Talk2Nature

World atlas: explore places and history →

THE KNOWLEDGE GRAPH · FIRST FOUNDATION

See the connections.
Keep the evidence.

Every note should lead back to a source. Every interpretation should keep its limits.

785 catalog records758 explicit citation linksNo recordings or private records

108 evidence notes · Open a note to follow its sources.

Japanese-tit call order: evidence and interpretations

Direct publisher full text reviewed for the 2016 primary paper (Methods, Experiments 1–2, Results, figures and ethics) and 2022 follow-up (selected Methods, Results, ethics, data/code statements and rights). The linked 2022 Figshare v1 tables were checked against the current official API sizes and MD5 values, paired on trial metadata, and summarized with the project standard-library script; the deposited R scripts were read as text, not executed. For the 2017 paper, publisher-indexed primary text exposed the abstract, selected Results, STAR Methods, statistics and data statement; its direct ScienceDirect page returned HTTP 403. The 2017 Mendeley Data v1 landing metadata and record-level license were reviewed, but its files were not. For Schlenker et al. (2024), the main text (Sections I–VII) of the official HAL v1 author-accepted manuscript was reviewed; the formal appendices were inspected selectively, not recalculated. The separate Wiley supporting-information PDF returned HTTP 403 and was not reviewed. Beckers et al. (2024) Sections 1–4 were read in the full text available through PubMed Central; its conceptual critique is not a new animal experiment. For the adjacent Coye & Dutour (2025) study (source 268), publisher-indexed Methods, Results and Discussion passages and the University of Western Australia repository abstract/access record were reviewed; direct ScienceDirect retrieval returned HTTP 403, so the full article file, figures, supplement and complete statistical details were not reviewed. No inferential model was reproduced and no independent Talk2Nature specialist review has occurred.

Read our interpretation and limits →

This note cites:
Patterns inside sperm whale codas

Primary study summary in founding audit; full-text methods review pending.

Read our interpretation and limits →

This note cites:
A birth, a group, and changing vocal styles

Primary study summary in founding audit; full-text methods review pending.

Read our interpretation and limits →

This note cites:
NatureLM-audio: a bioacoustic model, not a nature translator

Reviewed the ICLR 2025 proceedings abstract, architecture/training sections, BEANS-Zero tasks and evaluation, Tables 3–6, the BirdNET/Perch comparison caveat, conclusion and risk discussion. Reviewed Sections 2–4 and limitations of the 11-page arXiv v2 follow-up identified by its authors as a NeurIPS 2025 workshop paper. Checked the official overview/Quick Start, the Hugging Face model and dataset metadata, and the official code/data repository inventories and license files described in the source records. No dataset files, code or model weights were acquired, no inference was run and no results were reproduced. On October 13, also reviewed the BEANS-Zero dataset card at its visible commit, its live file inventory, and the official evaluation-code repository at pinned main commit 31d4487ee6452ae6c31853d45fd38b7d4150372d, including the project metadata. No dataset shards, sample rows, code or dependencies were downloaded or executed.

Read our interpretation and limits →

This note cites:
BEANS-Next and ROOTS: a broad bioacoustics benchmark, not a translation dataset

Targeted review of the October 2026 primary preprint: abstract, task taxonomy, data construction, benchmark split description, reported results/limitations, risks and reproducibility statement; also the official project page and Hugging Face cards. Full appendices, individual data rows, source licenses, code and model artifacts were not audited or acquired. No benchmark was run.

Read our interpretation and limits →

This note cites:
BirdNET: identification before interpretation

Official repository and license/review documentation reviewed for the audit.

Read our interpretation and limits →

This note cites:
Perch: reusable representations of animal sound

Official parent card and exact perch_v2_cpu/1 variation page reviewed, along with the repository migration notice, Hoplite README and pinned model configuration. No weights downloaded and no inference run.

Read our interpretation and limits →

This note cites:
DolphinGemma and shared signals

Official announcement, including its field-program location paragraph, reviewed; not an independent replication or complete dataset audit.

Read our interpretation and limits →

This note cites:
What bark classification can tell us

Full seven-page ACL Anthology paper reviewed on 2026-10-13, including Dataset, segmentation/annotation, all four task sections, five results tables and Limitations. No data, code or weights were acquired or executed; no result was reproduced.

Read our interpretation and limits →

This note cites:
Giving parrots a choice to connect

Published paper pages 3–12 reviewed for participant screening/ethics, protocol, analysis, results and limitations. Page 9 was rendered and Table 2's P1–P9 differences recalculated from the displayed session counts. The current University of Glasgow deposit record and Crossref DOI relation metadata were checked. No raw-data reanalysis or replication.

Read our interpretation and limits →

This note cites:
Learned buttons and receiver responses

University study report and vendor description reviewed; primary paper review pending.

Read our interpretation and limits →

This note cites:
Plant acoustic emissions: detection is not communication

Read the main 2023 Cell article and its full STAR Methods, including the recording, preprocessing, model evaluation and results sections; checked the relevant figure legends and selected supporting captions. Read the 2024 Opinion's historical framing, mechanism discussion, receiver-evidence conclusion and open questions. Also checked the exact Dryad record, Dryad's live CC0 policy and the linked Zenodo v1 classifier record. The Cell publisher endpoint was inaccessible, so its linked full-paper copy was used. No audio, archive, code or model was downloaded or run; no result was independently reproduced.

Read our interpretation and limits →

This note cites:
When another organism responds

Full version-of-record text reviewed through the NCBI PMC copy and eLife PDF, including main methods, figures, public reviewer comments and author responses. The public Mendeley Data v1 inventory lists three folders with 32 files (24 CSV, four R, two MATLAB and two XLSX); its page and file panels were reviewed separately from the article. On October 24, one 51,160-byte data.xlsx file was acquired to ignored external storage and matched the Mendeley file-panel SHA-256. A standard-library aggregate-only audit matched all five main Table 1 repetition, paired-observation and egg-cluster totals. No spreadsheet formulas or deposited code were executed, no inferential model was fit, the other files and publisher ZIP remain uninspected, and no independent specialist review was obtained.

Read our interpretation and limits →

This note cites:
Fungal electricity needs careful controls

Methods discussion and critique reviewed for the audit; no independent replication performed.

Read our interpretation and limits →

This note cites:
Fungal electrical activity: a sensor study, not a decoded language

Reviewed the full PMC article text, including Methods, Results and Table 1, figures, Discussion, data statement and article license, plus the Mendeley Data v1 description, displayed license and visible Files section. Table 1 reports electrode-pair counts, but the reviewed main text does not establish their independent culture-level replication. Supplementary Document S1 (including Method S1), raw dataset files and code bytes were not inspected or executed; no result was reproduced.

Read our interpretation and limits →

This note cites:
Start with tools researchers already use

Official tool documentation reviewed for the audit.

Read our interpretation and limits →

This note cites:
A benchmark is only as useful as its labels

BEANS task configuration and BEANS-Next card, displayed metadata and license description reviewed. No full data audit or experiment.

Read our interpretation and limits →

This note cites:
Monk parakeet voiceprints test caller identity, not meaning

Reviewed the final-article abstract and targeted methods/figure text surfaced by the primary PMC index, plus the official pinned GitHub README and the existing hash-pinned 2020 annotation sample. Recomputed the sample's Git blob SHA-1 and SHA-256; the Git blob matches the pinned GitHub file. Rechecked Edmond's released v1 API, current Data Access & Usage guide, ZIP-viewer FAQ, Terms of Use and official July 20, 2026 terms-update notice. No Edmond terms were accepted and no deposit archive/member was downloaded. Direct full-text retrieval was CAPTCHA-limited; the ANU repository PDF fetch timed out. The supplement, full statistical/code workflow and raw audio were not reviewed or run.

Read our interpretation and limits →

This note cites:
AVES and AVEX: reuse an encoder before training one

AVES abstract and official AVES/AVEX READMEs, including embedding examples and the AVEX license declaration. No model execution or checkpoint audit.

Read our interpretation and limits →

This note cites:
BIRB: test on the sounds your app will actually hear

Primary paper v2 introduction, benchmark overview and Table 1; repository BIRB entry. Detailed metrics and full experiments still need review.

Read our interpretation and limits →

This note cites:
DAS: build on an existing annotation workflow

Official introduction, repository README and bird tutorial through proofreading. The primary performance paper and current package were not evaluated.

Read our interpretation and limits →

This note cites:
Bat calls: context comes from observation

Primary Materials and Methods sections on housing, monitoring and classification tasks via PMC. Full statistics, supplementary analyses and raw data not reviewed.

Read our interpretation and limits →

This note cites:
animal2vec and MeerKAT: rare-call detection, not translation

Reviewed the 2026 published article's dataset methods, class definitions, annotation quality, evaluation design/results, discussion, ethics and data-availability statement; the official animal2vec repository and README; and the article's separate dataset, code and checkpoint license statements. The article supplement, deposited files/manifests, archived code contents and model weights were not inspected or acquired. No published result was reproduced.

Read our interpretation and limits →

This note cites:
Elephant calls: pair promising findings with their critiques

Selected discussion and playback methods in the author-organization PDF; Dryad landing page. The 2026 reanalysis was located through a publisher excerpt and related-work listing, not read in full.

Read our interpretation and limits →

This note cites:
Put the sound back into its scene

SuperAnimal abstract, introduction and dataset sections reviewed; SLEAP publisher summary and SimBA paper index and official project description reviewed. No code or model execution.

Read our interpretation and limits →

This note cites:
Plants have channels beyond sound

Primary abstract, introduction and selected results/methods passages reviewed; supplementary data not reanalysed.

Read our interpretation and limits →

This note cites:
Fungal electrical signals travel; communication remains untested

Full text of the 2026 primary paper and its Methods, limitations and data statement reviewed, alongside the official cited Zenodo record. The paper's eight channels come from one specimen; the 2021 record lists one 413.2 MB archive but does not map the analyzed 15-day series to a file. The dataset license is not displayed. No files were downloaded, no analysis was reproduced and no independent specialist review was received.

Read our interpretation and limits →

This note cites:
Do fungi have a language? What electrical recordings show

Reviewed the full arXiv HTML text of Adamatzky's 2021 preprint, including recording methods, event detection, grouping rules, results and discussion; checked the published 2022 Royal Society article metadata and institutional article-rights record. Reviewed the 2024 critique's publisher abstract and visible selected sections, not its full text. Reviewed the full 2025 FEMS review sections on fungal electrophysiology, macroscopic mushroom recordings, the authors' method-replication attempt, caveats and future methods, plus its article license. Inspected the official Zenodo v1 record and file metadata only. No files were downloaded, no analysis was reproduced, and no independent specialist review was received.

Read our interpretation and limits →

This note cites:
Ice-crystal claims: why beautiful images need stronger tests

Reviewed the full 2006 four-page article, full 2008 Journal of Scientific Exploration paper, Douglas Stokes’s 2009 letter (p. 83), and Kokubo, Nemoto and Kawano’s 2017 technical note (pp. 73–86). Checked the reported designs, outcomes and experimental units. The 2006 full text was read from an author-hosted copy; its indexed PubMed record remains linked. Raw photographs, original analysis files/code and a full systematic literature search were not reviewed; no independent specialist review was obtained.

Read our interpretation and limits →

This note cites:
Parrot owners can contribute; species still matter

Full main article text, including methods, results, discussion, data-availability and rights statements, reviewed through the Europe PMC full-text XML. Official Figshare v4 API metadata and its single-file manifest were checked. The supplementary questionnaire (Appendix 1), workbook, and raw responses were not inspected or acquired.

Read our interpretation and limits →

This note cites:
When companion parrots use names: what a survey can and cannot show

PLOS ONE article HTML reviewed for abstract, methods, results and discussion; the official correction notice was checked. Supporting information, raw survey data and figures were not downloaded or inspected, and no independent reanalysis was performed.

Read our interpretation and limits →

This note cites:
Companion-parrot S2: reported context is not a call label

Reviewed the primary PLOS paper's methods and S2 description and the NLM Dataset Catalog record, which links the S2 DOI to PLOS Figshare item 32046233 and reports CC BY 4.0. The direct Figshare page/API were not retrievable in this review. No workbook bytes, schema or cell contents were inspected; no reanalysis was performed.

Read our interpretation and limits →

This note cites:
Budgerigar vocal signatures: caller identity, not call meaning

Reviewed the full Current Biology article text through NCBI PMC E-utilities (PMCID PMC10799565), including call-elicitation, VAE/SVM methods, results, discussion, animal-care and data/code statements. Rechecked the version-1 Mendeley landing-page description and API documentation against the prior checksum-pinned one-file inspection. No new data files were acquired or parsed; no code/model was run, no reanalysis or independent review was performed.

Read our interpretation and limits →

This note cites:
Cats: useful labels, limited reuse

Version 1.0.2 record, API rights and naming conventions; paper abstract only.

Read our interpretation and limits →

This note cites:
Pig calls: inspect what the labels actually mean

Primary indexed abstract/results excerpts and repository description/API; full methods not reviewed.

Read our interpretation and limits →

This note cites:
Frog choruses: several callers, one recording

Full data-availability and split sections and selected results in the paper; official Zenodo API metadata; README, exact pinned MIT license and evaluator source read as text. No replication or code execution.

Read our interpretation and limits →

This note cites:
Before AI: learning from the honeybee dance

Selected lecture passages on dance distance/direction and the 1949 orientation trial reviewed; PDF pages 9–10 visually checked. The referenced original experiments and subsequent literature have not all been reviewed.

Read our interpretation and limits →

This note cites:
Fish hear a habitat, not a translated sentence

Abstract, selected results/discussion and Methods passages on study site, acoustic treatment and controls reviewed. Full supplements, analysis code and independent replication not reviewed.

Read our interpretation and limits →

This note cites:
A shared task can make a shared signal useful

Primary abstract and indexed passages from the university’s research-team account reviewed. Full paper methods, supplementary material and data were not obtained; no independent replication performed.

Read our interpretation and limits →

This note cites:
Context helps vervets interpret alarm calls

Wiley's 1980 publisher abstract and German summary, the ScienceDirect publisher abstract for the 1980b companion paper, the 1980a Science PubMed abstract, selected full-text sections of Fischer (2020, pp. 108–114), and the full open-access 2023 Springer primary article were reviewed. The 2023 paper's methods, measured responses, sample scope, rights statement and data-availability statement were checked. On 2026-10-25, an unauthenticated GET to the canonical OSF API v2 node endpoint succeeded: it reports the linked project as public, gives its title and creation/modification dates, and returns a null project-level license relationship. The publisher-provided view-only URL separately returned HTTP 403. No OSF file listing, contents or component-level terms were reviewed.

Read our interpretation and limits →

This note cites:
A shared song can change across years

Indexed 1985 publisher abstract and 2017 review abstract; full methods, spectrograms and original audio not reviewed.

Read our interpretation and limits →

This note cites:
Sometimes the signal is something shown

Publisher abstract and figure captions reviewed; full methods and supplementary videos not obtained.

Read our interpretation and limits →

This note cites:
Who leaves matters to the wolves who stay

Selected results, discussion and main methods via Europe PMC XML; supplementary procedures, data and statistical replication not reviewed.

Read our interpretation and limits →

This note cites:
Two hunters, different skills, a visible signal

Selected results, discussion and Materials and Methods reviewed; videos and raw records not reviewed.

Read our interpretation and limits →

This note cites:
A familiar whistle can address an individual

Indexed primary abstract, discussion and methods location passage reviewed. Complete methods, supplements and raw recordings not obtained.

Read our interpretation and limits →

This note cites:
What parrot touchscreen research can—and cannot—tell us

On October 14, reviewed the full article text from the 16-page published PDF hosted by the University of Glasgow, from the abstract through the conclusion, including the caregiver survey/pilot, participant criteria, ethics, protocol, data cleaning, results, limitations and design guidelines. Figures were not visually reanalyzed; no raw data, code, supplements or independent analysis were available or inspected.

Read our interpretation and limits →

This note cites:
Budgerigar vocal and social records are not event-aligned

Reviewed the primary article's animal housing, experimental outline, audio recording and video recording Methods through PMC-indexed full text and a coauthor-hosted article PDF; the official Dryad version history, file inventory and README data dictionary; Dryad reuse guidance; and the official NMSU facility locator. Direct PMC page access encountered CAPTCHA. No Dryad files, raw audio, video, supplement or code were downloaded or executed. The README describes the interobserver workbook, but it was not inspected and agreement was not recalculated.

Read our interpretation and limits →

This note cites:
MammAlps aligns animal sound with video-annotated behavior

Targeted review of arXiv v2 sections 3.1–3.4, 5, 7.1 and 9.1–9.2; exact Zenodo v3 metadata, README, file inventory and dataset license; official CVPR Open Access indexed excerpts from the main paper and supplementary section A.1; official GitHub README, evaluation interface and LICENSE.md. The direct CVF paper and supplement fetch returned 403, so only targeted indexed excerpts were reviewed for the published version. Other supplementary methods and archive contents were not reviewed; no data or code were downloaded or run.

Read our interpretation and limits →

This note cites:
Parrotlet babbling is vocal development, not call translation

Targeted PMC-indexed excerpts of the primary paper's abstract, conclusion and data-access statement; the exact Dryad dataset landing page, public API dataset/version/file records (v6, version record 180872), current End User Terms and Publication Policies; and an indexed excerpt from the official KBA location factsheet. Dryad's published policy states accepted datasets are made public under CC0, with no separate exception shown on this record. API file metadata supplied exact IDs, sizes and SHA-256 digests. An unauthenticated byte-range request to the API download route returned 401; the web file stream returned HTML rather than the file. No paper PDF, supplement, dataset file or audio was downloaded or fully reviewed.

Read our interpretation and limits →

This note cites:
Natural-language prompts can retrieve annotated wildlife videos

Targeted review of arXiv v1 sections 1, 3.1–3.2, 4–7 and supplement S1.1–S1.3; official EPFL project page; pinned Hugging Face dataset card; official GitHub README, pyproject, SALMA README, repository root inventory and checkpoint issue. No dataset, code, article or model weights were acquired; no code or result was run.

Read our interpretation and limits →

This note cites:
Fruit-bat calls and context: a rich dataset, not a translation key

Reviewed the 2017 Scientific Data article's abstract, methods, data records, Tables 3–4, technical validation, usage and rights sections; the 2016 Scientific Reports results and classification methods; the 2023 iScience abstract, dataset description, sequence-HMM methods/results, discussion and data statement; the Mendeley Data version 3 description/license and official version 1/2 and 2/3 comparisons; the Mendeley public file-list API documentation and one anonymous version 3 request (HTTP 401; documentation specifies OAuth2); the NLM catalog entry; and direct public Figshare API metadata for archive item 4556074, annotation item 4555903, FileInfo item 4555897 and Metadata.pdf item 4555900. Acquired only the 3.29 MB annotation table with verified size/MD5 for aggregate-only schema/category-count comparison, plus the 367,257-byte metadata PDF for visual review of all five pages; its source API supplies no MD5, so the receipt's SHA-256 is local provenance only. No audio/video, FileInfo.csv, code or model weights were inspected, and no published model result was reproduced.

Read our interpretation and limits →

This note cites:
Wild parrot duet structure: sequence patterns, not decoded meaning

Reviewed the publisher article HTML for its abstract, sites and recordings, call categorization, acoustic analyses, syntax methods and results, discussion, data-availability statement and supplementary-file inventory. Reviewed the full Dryad dataset landing page and README description and Dryad’s current reuse-rights guide. The workbook and publisher supplements were not downloaded or inspected; no published analysis was reproduced.

Read our interpretation and limits →

This note cites:
A parrot can imitate a call to address another bird. What the experiment shows.

Reviewed the full primary PLOS article text for the abstract, introduction, Materials and Methods, Results, Discussion, Ethics Statement and article license. No raw recordings, associated dataset, supplementary material or analysis code were acquired or inspected; no reanalysis was performed.

Read our interpretation and limits →

This note cites:
A 2024 model study predicts bat context—with important limits

Reviewed all five pages of the official EUSIPCO 2024 proceedings paper and the official arXiv v1 record/manuscript, focusing on dataset filtering, evaluation protocol, model setup, metrics, class errors, data source and license. Also reviewed the first author’s linked public GitHub profile and the pinned Earth Species Library snapshot at commit 6382c52bf553569528e0a790efe16e5ae42f0094: root inventory, bat README and displayed task/split/license sections from notebooks 01, 02, 99a and 99b. No code, split files, raw media, model weights or supplements were downloaded or run. On 2026-10-23, the first author’s live GitHub owner-repository API again listed 13 public repositories (newest update 2025-10-20), and a GitHub repository search for the exact paper title returned no match. This remains a bounded search, not proof that code or splits are unavailable elsewhere. A Zenodo API search was inaccessible from this environment and is not treated as a negative result. The acquired five-page Figshare Metadata.pdf item 4555900 was visually reviewed across Tables 1–4. It lists 20 treatment periods with date ranges, group composition and channels; subject IDs, mother/date-of-birth fields; and an annotation schema with both emitter and addressee. The already-acquired 91,080-row Annotations.csv item was rechecked with an aggregate-only participant-pair audit. No raw annotation rows or IDs were exported. The 31.6 MB FileInfo.csv item and the paper’s exact fold files remain unacquired/unavailable. On 2026-10-23, the checksum-verified Annotations.csv was filtered using the paper-described context, emitter and duration rules. The aggregate-only script matches the reported 35,074-row count and summarizes file-plus-emitter and file-plus-all-known-participant group topology; it exports no row or identity values. No audio, FileInfo.csv, paper fold files, model code or model weights were acquired or run. On 2026-10-23, the official St Andrews group page was reviewed for bat vocal-communication expertise, and the TAU profile plus Nature 2025 author list were checked solely to screen fit and disclose a separate Vernes–Yovel coauthorship. No reviewer was contacted or selected. On October 24, 2026, the official Figshare API and exact CC0-declared annotation file size/MD5 were rechecked, then the aggregate audit was regenerated. The new caller-support diagnostic excludes 122 mixed-caller files and retains 99.26% of the filtered annotations; it creates no split.

Read our interpretation and limits →

This note cites:
BioCPPNet separates overlapping calls—it does not decode them

Reviewed the article abstract, Methods, Data availability, Code availability, Results, Discussion and CC BY 4.0 statement, plus the linked repository README and visible root inventory at commit 7092851cbd9a32116e6453583823d5334ab11f1b. No code, recordings, generated mixtures, supplementary media or trained weights were acquired or executed.

Read our interpretation and limits →

This note cites:
Do animal calls function as names? A proposed test framework

Reviewed the official PubMed record and ScienceDirect listing for article metadata, abstract, publisher highlights, Open Access/Creative Commons status and copyright information. The direct publisher page returned HTTP 403. Full text, references, supplementary files and underlying studies were not reviewed; no data or analyses were acquired or reproduced.

Read our interpretation and limits →

This note cites:
Dupont's lark songs and calls: an annotated vocal-type dataset

Reviewed the full publisher version of record, including Abstract; Methods (study area, recording, annotation and vocalization identification); Data Record; Technical Validation; Usage Notes; Data Availability; Code Availability; and Rights and Permissions. Also reviewed the official Zenodo API response: concept DOI 17264916 returned HTTP 302 to version record 17264917, whose JSON declares conceptdoi 17264916, version DOI 17264917, record-level CC BY 4.0 and eight files with provider-supplied sizes and MD5 values. Reviewed Datos.gob.es metadata, the full 4,169-byte Digital.CSIC README and archive HEAD metadata. No dataset archive, code, article file or figure was saved or executed; provider checksums were not locally verified.

Read our interpretation and limits →

This note cites:
Can trees talk through fungi? What mycorrhizal networks show

Reviewed the Nature publisher record and surfaced article passages plus its author correction; the New Phytologist summary and selected discussion; and the full publisher-hosted Frontiers Opinion, including its article type and author statements. Direct rendering of the Nature article required publisher authentication in this review. The underlying studies, supplementary files and data were not reanalyzed, and no specialist review was obtained.

Read our interpretation and limits →

This note cites:
Macaw calls change with age and population. That is not a translation.

Reviewed the publisher abstract and exposed section snippets; the publisher page itself was not fully accessible. Also reviewed the Dryad record, abstract, README descriptions and file inventory, and Dryad’s current data-reuse policy. The README says the RDS selection objects combine call selections with corresponding sound files through warbleR, but their contents were not inspected. No article, data, audio, code or supplementary bytes were acquired or executed.

Read our interpretation and limits →

This note cites:
Speech models carry animal vocal patterns—not animal meanings

Reviewed the complete arXiv v1 HTML, including dataset descriptions, encoder families, caller-identification and call-type tasks, fold construction, result tables, the session-independent sensitivity analysis and conclusion. Confirmed the submitted version/date and CC BY 4.0 manuscript license on the arXiv record. No paper-linked code or data release was identified in the reviewed record/full text; this is not an exhaustive search. No audio, dataset files, code or weights were acquired or run. The arXiv record does not establish publication or peer-review status.

Read our interpretation and limits →

This note cites:
Monk parakeet calls carry individual identity information

Reviewed the PLOS article’s sampling, methods, results, discussion, ethics, and linked-resource sections. Refreshed the Figshare v2 API and streamed its README.md file 41233869 for review; it was not retained. Resolved the GitHub v1.0.0 tag to commit 1f7031dc996c79914ccf5d042181f947793fb459 and read the site-scale selection R Markdown file at that revision without retaining or executing it. The code explains the 14-row selection difference. The PLOS supplement, remaining code/dependency workflow, selection CSV, and audio were not reviewed. No dataset files or audio were acquired; no code was executed.

Read our interpretation and limits →

This note cites:
Whombat: learn from an existing bioacoustic annotation workbench

Reviewed the complete primary article through the publisher/UCL open-access record and the pinned Whombat README/license and PyPI release metadata. Read the official import, dataset, annotation-project, evaluation and installation guides. Verified the temporary Whombat 0.9.0 wheel SHA-256 against PyPI and inspected the separate Whombat GPL v3 and soundevent MIT licenses. In an isolated Python 3.12.5 environment on macOS 26.4, the exact current Listen export passed Whombat's dataset-then-project UI import. The app created one recording, two TimeInterval annotations, two notes, one task and four project tags; checksum, sample rate and duration matched, the tag table showed both sound events, and both imported creator IDs remained null. Core resolved versions were FastAPI 0.112.2, fastapi-users 12.1.3, Starlette 0.38.6 and SQLAlchemy 2.0.54. The test used only an invented WAV and a temporary SQLite database; the wheel, environment and test outputs were removed. No animal audio, participant record or model weight was tested.

Read our interpretation and limits →

This note cites:
A shared bioacoustic dataset loader, with rights kept separate

Read the public alp-data README and project configuration and reviewed the exact PyPI 1.11.1 release page and MIT license declaration. No package, dependencies or datasets were downloaded, installed or run. No individual dataset entry was cleared for use.

Read our interpretation and limits →

This note cites:
ANIMAL-SPOT: parrot call detection is not translation

Read the full publisher article text, including the target/noise and call-type methods, preprocessing, split policy, results, discussion and ethics/data-collection description, alongside Tables 1 and 3 from the earlier targeted full-text review. The paper states that samples from one recording stay in only one split, with a default 70/15/15 train/validation/test split; it does not establish individual-animal holdout. It reports mean 97.9% test-set accuracy (range 94.5–99.8%) and, separately, 95.9% average AUC (range 91.7–99.1%) across 33 additional unseen animal-specific recordings. Monk-parakeet call-type accuracy is 92.7% on human-labeled test samples and 88.4% on detector-selected segments. Re-read Edmond V2.0 API metadata on October 23, 2026: file restricted=false, fileAccessRequest=true, dataset-level CC BY 4.0 and the same provider MD5. The official Dataverse 6.11 API documentation says access-request settings are ignored when a file is unrestricted. Edmond's own download guide separately requires acceptance of its terms and dataset license; the current Terms' MPS-associated service wording still does not explicitly clarify its application to external downloaders of existing public datasets. The supplement's split tables, exact V2 archive manifest and individual/session recurrence remain unreviewed. No data, code or model was acquired or run; no terms were accepted and metrics were not reproduced.

Read our interpretation and limits →

This note cites:
A community directory for discovering bioacoustic datasets

Reviewed the README and root inventory at commit c412c5c3fd65552d17e492dac35f2be2c30349a4. The README describes a community list of online bioacoustic datasets intended for machine-learning use, its metadata fields, and how people can submit corrections or additions. GitHub API metadata reports no recognized software license, and the pinned root inventory has no LICENSE file. No dataset entry, linked file or model was acquired, installed, copied or run.

Read our interpretation and limits →

This note cites:
CoarseSoundNet: soundscape context, not animal language

Reviewed the full arXiv v2 HTML: data composition, model comparison, transfer to BEsound, silence handling, threshold/duration post-processing, error analysis and ecological-index case study. Reviewed the pinned GitHub repository/README/license, Hugging Face revision and file/rights metadata, and the Zenodo record plus read-only API metadata. No code, model weights, dataset archive, audio, figures or supplementary files were downloaded or run.

Read our interpretation and limits →

This note cites:
WhAM: acoustic style transfer is not whale translation

Read the full arXiv v1 main text from abstract through limitations, and reviewed targeted Appendix E sections on training data, compute, downstream classification and expert perceptual evaluation. Checked the official repository README/license and pinned source revision, the DSWP dataset card and its current repository revision, and the Zenodo model-weight record. No code, audio, datasets, model weights, paper figures or supplementary files were downloaded or run; Appendix D and artifact contents were not fully audited.

Read our interpretation and limits →

This note cites:
Biodiversity standards can carry media and provenance, not animal meaning

Reviewed the Darwin Core Data Package guide's normative requirements and versioned profile, including the requirement to use official per-table field descriptors; the conceptual model sections on Occurrence, Identification and Media; selected current Audubon Core terms covering media rights, access points, hashes and reviewer metadata; GitHub state/comments for issue #759; and the official rs.tdwg.org repository tree. The versioned profile and table-schema URLs linked by the guide returned HTTP 404 in this environment. The repository's profile and event-schema paths trace to commit e0cc2509953eeed24ad312b6e91e7bfcadd37a3b, titled 'Initial pre-review DwC-DP schemas'; the repository source is unversioned and is not used as the released 1.0 profile. No package or field mapping was generated or validated.

Read our interpretation and limits →

This note cites:
How great tits respond to a distant species’ mobbing calls

Reviewed the publisher-indexed Methods, Results and Discussion passages and the University of Western Australia repository abstract and publication/license metadata. The publisher states 96 wild great tits, six playback treatments with 16 birds each, and a single playback per bird. The indexed text exposes stimulus construction and behavioral outcomes; direct ScienceDirect retrieval returned HTTP 403, so figures, supplements, full statistical details, and article-level context were not reviewed.

Read our interpretation and limits →

This note cites:
Common endophyte networks: resource transfer is not communication

Full publisher article reviewed: abstract, introduction, results, materials and methods, statistical analysis, data-availability statement, limitations and rights. Official GitLab project metadata, current main commit, root tree and per-file metadata were checked; no dataset bytes, R code, README or supplementary file was retrieved or run.

Read our interpretation and limits →

This note cites:
Lilac-crowned Amazon calls vary with context; meaning remains a question

Reviewed the full publisher article for abstract, field methods, context definitions, acoustic measurements, statistical analyses, results, discussion, data availability, ethics and rights; reviewed the additional-file inventory. The research viewer failed or timed out on selected binary supplements. No article assets, audio, data or code were downloaded, opened, executed or redistributed; no independent reanalysis or specialist review was performed.

Read our interpretation and limits →

This note cites:
Phones can record wildlife, but they are not interchangeable sensors

For Jäckel et al. (2021), reviewed the PLOS paper's methods, citizen-recording quality results, device playback comparison, discussion, rights and Zenodo data statement; the cited Zenodo deposit remains unreviewed. For Gili et al. (2024), reviewed the full publisher article and the March 13, 2024 Dryad record, visible file list and embedded README structure, as well as Dryad's CC0 and responsible-sharing policies. The separate README download returned HTTP 403. No audio archive, article supplement or code was acquired or reproduced, and no independent specialist review was completed.

Read our interpretation and limits →

This note cites:
Strong bird-call labels can test detectors, not communication

Reviewed arXiv v2 methods, annotation design, dataset schema, technical validation, project-split evaluation, discussion and license; the exact Zenodo v5 API manifest; and pinned repository source. All 12 provider-reported filenames, byte sizes and MD5 strings match this catalog; no dataset file bytes were hashed. A 4,096-byte HTTP 206 range of metadata_v2.csv was streamed transiently and only its header line inspected; it names location_id, recorder_id, date_recorded, latitude and longitude. Raw GitHub source files were fetched read-only into a transient filter for static review; none were saved to the project, installed or executed. No data-row values were reviewed or saved; no annotation rows, manifest rows, audio or checkpoints were inspected. No independent reproduction or specialist review occurred.

Read our interpretation and limits →

This note cites:
OpenSoundscape: bioacoustic tools, not animal translation

Reviewed the paper abstract, software overview, annotation and localization functionality, examples, and stated scope; the official v0.13.2 release and tagged MIT license; the official annotation tutorial; and the Dryad dataset landing page and repository-wide license policy. The article describes v0.9.1, the repository lists v0.13.2 as its latest release, while the stable annotation tutorial still pins 0.13.1. No package, code, model, recording, archive, or annotation row was downloaded, installed, or run. The publisher marks the article Open Access, but its exact reuse license was not visible in the reviewed page. No independent software or biological review occurred.

Read our interpretation and limits →

This note cites:
HyenaSET: a strong sound-event benchmark, not a communication test

Read the bioRxiv v2 abstract and targeted data-collection, annotation, split, benchmark and data/code-availability sections. Examined official Edmond and DataCite metadata for version 1, release date, archive size, provider checksum and dataset terms. The paper is a preprint, not peer-reviewed. No audio, labels, code or model weights were downloaded, installed or run; no data rows, coordinates or individual timestamps were inspected. No independent review or reproduction occurred.

Read our interpretation and limits →

This note cites:
CowRumIA: a cattle-audio app and a test-set caution

Read the pinned CowRumIA README and LICENSE, statically inspected model/train_model.py around its split, callbacks and final evaluation, and reviewed the Zenodo v1 record description, file listing, dates and Rights section. The repository snapshot was pinned to commit 44fdc5e5cf3f080457211bb87eb8cb2fc6b099ac. No dataset or model files were downloaded, no code was run, no app was built, no thesis full text was reviewed, and no independent specialist review or reproduction occurred.

Read our interpretation and limits →

This note cites:
Playback can shape the call you measure

PubMed citation/abstract and selected PMC author-manuscript sections surfaced by the article index—subjects, apparatus, stimuli, procedures, analyses, and results/discussion—reviewed 2026-10-14. The listed supplement, raw data, and code were not inspected; no reanalysis, replication, or independent review. Reuse terms were not verified.

Read our interpretation and limits →

This note cites:
A camera can decode a known bee signal—with calibration

Targeted PLOS ONE full-text sections reviewed: abstract, communication background, recording setup, target features, software modules, validation/results, discussion and license. The official GitHub README, API metadata and complete recursive tree were reviewed at pinned commit 3aaa5099aac698bd08df1d55f4cd5183d32f7e7d. Selected setup, validation, runtime and rights sections of the related 2015 Frontiers system paper were reviewed. No data, video, code, weights or supporting files were downloaded or executed; no reproduction or independent review.

Read our interpretation and limits →

This note cites:
Barkopedia: context labels do not show call meaning

Reviewed the official Barkopedia README and MIT LICENSE at GitHub commit ef8da9155e101cb76d4ee260bebb4d4fde4c2a1d; opened the linked task Spaces; reviewed the Activity and Environment dataset card and Hugging Face revision/tree at 60a721a7f9477634971f6b3df29dc2a5e56994d8. The README says source audio was collected from YouTube and Reddit and lists eight tasks. Only the Activity and Environment task card was examined; no academic challenge-results paper or final leaderboard report was reviewed. No archive, CSV, audio, code or model was downloaded or executed, and no independent reproduction or specialist review occurred. On October 14, also inspected the exact Hugging Face Task 6 card revision and its README diff: the former Barkopedia emotion card was rewritten as EmotionalCanines and changed its license metadata, but this does not establish dataset-file continuity. No Task 6 audio, labels or archives were inspected; the Activity and Environment card remains the only Barkopedia task-specific dataset card reviewed.

Read our interpretation and limits →

This note cites:
Amazon parrot dialects across 22 years: cultural change, not call meaning

Reviewed the article abstract and targeted Methods/Results passages surfaced through the PMC search index; direct publisher retrieval returned HTTP 403 and the PMC page presented a bot check, so the full article and supplements were not reviewed. Reviewed the Dryad v6 landing-page data dictionary and rights statement, current Dryad reuse guide, TLS-verified API version/file manifest, and the GitHub README, license metadata, pinned tree and setup/input sections of four R Markdown analyses. No dataset bytes were acquired; no code or model was executed.

Read our interpretation and limits →

This note cites:
EmotionalCanines: video-derived labels are not a dog-reported emotion

Read the author-hosted eight-page ACM Multimedia paper sections on dataset generation, annotation, reliability, dog-separated splits, baselines, results and limitations. Reviewed the exact Hugging Face card commit c0617b15562882b9eee4cc1bcb8e6aa80e51bd34 and its README diff, plus the pinned GitHub README/LICENSE at 6b73c5fce55fc132e97db6a76899f8594a955f59. The paper reports 1,400 bark sequences (700 per breed; 600 train and 100 test per breed) from 199 dogs, with dog identities separated between train and test. Twelve annotators contributed; each sequence received three labels, and a 100-sample pilot reported Fleiss kappa 0.30 for arousal and 0.42 for valence. No recordings, labels, code or model weights were downloaded or run; no metrics were reproduced.

Read our interpretation and limits →

This note cites:
El Silencio helps detect sound across taxa, not decode meaning

Queried official Zenodo API metadata for labeled record 17064353 and unlabeled record 17122738, including publication dates, record-level licenses, file counts, byte totals and provider MD5 values. Read the dataset repository README and MIT LICENSE at commit 5e2f7826c908fec3018740068bbdeda6c6977f35; reviewed all extractable Spanish text on pages 2-18 of its pinned protocol and the full 11-page CEUR overview paper. Local review copies of only the paper and protocol are excluded from the public site. No dataset audio, label table, code or model weights were acquired or run; no baseline was reproduced.

Read our interpretation and limits →

This note cites:
CEB can benchmark bird sound detection, not what a call means

Reviewed the official Zenodo v1.1 landing page, published file inventory, component-specific rights statement and schema.json browser preview. The v1.1 notes identify annotation and taxonomy corrections; the prior v1 record 20762099 is superseded. Re-read the official Zenodo API inventory on October 23, 2026: train_soundscape.csv is 8,559,233 bytes (provider MD5 55b5e3f99566c5bb3ba4fb5b833424da) and test_soundscape.csv is 2,401,049 bytes (provider MD5 f572155ae41714e17b218e95d97b1ce9). These checksums are upstream metadata, not locally verified. No CSV rows, audio, code or weights were downloaded, played, parsed or executed; no detector was evaluated.

Read our interpretation and limits →

This note cites:
BirdSet: a large bird-sound benchmark, not animal translation

Reviewed the primary arXiv abstract and version history, the current BirdSet dataset card and its configuration/metadata/license descriptions, and the complete maintainer clarification thread on mixed per-record licenses. No paper body, audio, labels, code or weights were acquired or run; no independent specialist reviewed this note.

Read our interpretation and limits →

This note cites:
WABAD maps bird calls at scale; it does not map meaning

Reviewed the publisher abstract and data-availability statement, plus the official current Zenodo v4 record, version history and file inventory. The full article body, supporting-information PDF, archive contents, metadata schema, labels, evaluation split and code were not inspected; no data or models were acquired or run.

Read our interpretation and limits →

This note cites:
OpenWhistle: a strong dolphin sound benchmark, not a translator

Reviewed the arXiv v1 paper data-collection, annotation, benchmark, evaluation, limitations and ethics sections; revision-pinned Hugging Face pretraining and classification dataset cards and file inventories; the separate revision-pinned model card; and the pinned official OpenWhistle README (source 411) and LICENSE-DATA (source 412) at commit ad715bb3cf24c36f85b3c3c8159077548134a39e. The repository README applies CC BY 4.0 to its four named dataset releases and LICENSE-DATA grants those terms; the dataset-card API license fields remain null. No audio, labels, code or weights were acquired or run. No independent specialist reviewed this note.

Read our interpretation and limits →

This note cites:
Parrot social ties track call variation, not call meaning

Reviewed the published article abstract and targeted methods, results, discussion, data-access, ethics and rights passages using the primary full-text record and institutional copy; checked the 2023 voiceprint paper for sample overlap; reviewed the Zenodo record and the GitHub README/tree at a pinned revision. On October 24, rechecked Methods section 2.1 and confirmed the Barcelona study region and field dates for the atlas. The archive, raw recordings, code and model outputs were not acquired or run.

Read our interpretation and limits →

This note cites:
iNatSounds scales species recognition, not animal meaning

Reviewed the paper abstract and targeted dataset-construction, evaluation, weak-label and assets/consent passages; the complete paper was not exhaustively reviewed. Reviewed the complete 2024 dataset README at commit bd52ecde5d9d7434f6167bbac823d152c3823ccd, including class counts, single-species file labels, location/license fields, ToU and provider-listed archive sizes; the current iNaturalist media-license FAQ; the pinned code repository README and MIT SPDX metadata; and the author-linked model folder pointer. No audio, annotation archive, source code, dependencies or weights were downloaded or run.

Read our interpretation and limits →

This note cites:
V33DA: a zebra-finch caller benchmark, not a translation model

Reviewed the pinned V33DA dataset card, code/sample README, split-preparation script, training entry point and read-only Git-tree inventory. Fetched the sample README.md, metadata.json and split_summary.json over verified TLS; computed in-memory SHA-256 values matched the pinned Git blob IDs. See the manifest for checksums. No text copies were retained. No Parquet, audio, accelerometer, video, calibration, executable code or model bytes were acquired or run. Maintainer metadata review only; no specialist review. A follow-up static import/tree check confirmed that scripts/train.py imports v33da.data modules absent from the pinned public repository tree; the split-preparation script materializes only session and held-out-experiment assignments. No code was run.

Read our interpretation and limits →

This note cites:
A large forest-bird soundscape set can test detection, not conversation

Reviewed the publisher title, abstract, publication notice and rights statement; the full article and supplement remain unreviewed. Read the official Zenodo v1/v2 record descriptions and complete v2 API manifest. Acquired only ReadMe.txt, site_metadata.csv, recording_metadata.csv and species_metadata.csv from v2. Verified provider MD5 declarations against bytes and recorded local SHA-256 values in an ignored receipt. The aggregate script emits no raw site IDs or coordinates. No recordings, annotation rows, code or models were acquired or run.

Read our interpretation and limits →

This note cites:
The Wytham great-tit dataset maps song structure, not meaning

Reviewed the 2024 publisher article and author dataset documentation for collection design, labels and package format; Oxford Research Archive rights record; setup-code license and pinned revision; and the full 2025 Current Biology article including STAR Methods on feature extraction, similarity, clustering and validation. The 2025 methods describe a ViT metric-learning model trained on song-type labels from this corpus, 384-dimensional embeddings calculated for all 109,963 songs, and downstream repertoire-similarity/cultural analyses; the article reports 0.87 validation accuracy for a separate ResNet50 procedure on a random subset. The reviewed method text does not establish an individual-held-out test for that validation. The complete 2026-10-24 OSF root traversal lists 23 current file entries totaling 12,184,678,510 bytes: 14 data files and nine documentation files. An earlier same-day 14-entry traversal had listed only the data/ folder and omitted the separate docs/ folder; it is superseded. All 23 normalized paths and byte sizes match the 2026-10-23 snapshot. The saved 2026-10-23 SHA-256 for docs/metadata/main.md is malformed at 63 characters. On 2026-10-25, the 6,426-byte file was retrieved solely for integrity verification; its SHA-256 and MD5 match the current OSF API. Its content was not read or redistributed. The OSF node license relationship declares CC BY 4.0. The author nestbox documentation specifies British National Grid X/Y coordinates and says these data are also included in main.csv. No coordinate values, audio/annotation or analysis-code payloads were accessed; the one documentation file was not read or redistributed; no model was run and no independent specialist reviewed this Talk2Nature note.

Read our interpretation and limits →

This note cites:
Bat auditory cortex anticipates call patterns, not their meaning

Reviewed the full main article through PMC, including its question, sample, acoustic and neural recording methods, category definitions, results, discussion, data/code availability and article license. Read the GIN DOI landing-page metadata only. The 5.8 GiB archive, repository file list, README, scripts, supplementary information and peer-review file were not inspected or acquired. No independent specialist reviewed this Talk2Nature assessment.

Read our interpretation and limits →

This note cites:
A year of budgerigar vocal activity is not a conversation

Reviewed the publisher early-access record and abstract only. The page identifies the article as an early-access version and says complete PDF/HTML/XML versions will follow. A direct full-page request was rate-limited. Detailed methods, model validation, ethics, data availability, code, weights and supplements were not reviewed; no files were acquired or run.

Read our interpretation and limits →

This note cites:
V33DA++ adds overlapping calls and context, not meaning

Reviewed the V33DA++ Hugging Face dataset card at immutable revision bb31f73e2191278dd4965af01502c42380cd597d, including bucket counts, data-field definitions, suggested tasks and rights notice; checked the provider API revision/date/license metadata. Rechecked the linked GitHub repository's current main revision e988b985e38909b40b2705cdad994e3d9307181e, its training imports and split-preparation script as public text, and the Git tree. No dataset files, media, Parquet, code or weights were acquired or run; no specialist review occurred.

Read our interpretation and limits →

This note cites:
Zebra-finches exchange calls with an interactive audio model

Reviewed the full bioRxiv v3 main text, including Results, Discussion, Methods 1.1–1.7, data-availability statement and posted license, plus official Zenodo landing-page/API metadata for records 23024019 and 23047646. A metadata-only refresh of the official bioRxiv v3 page on October 24 confirmed its current title and that the title and abstract were updated; the full text was not re-reviewed in this refresh. The code/model record lists one ZIP; the separate data record lists 11 files, including four raw example recordings and processed archives. No archive internals, supplemental analyses or media payloads were downloaded or inspected; no code, weights or data were run; no independent specialist reviewed this assessment.

Read our interpretation and limits →

This note cites:
Zebra finches categorize call types; translation remains a bigger question

Reviewed the full PMC-indexed article text, including the abstract, reported results, Figure 4 caption and data statement on October 24; the UC Berkeley institutional-copy record; the authors' public GitHub README/root listing; and the 2026 BirdAVES article. On October 25, refreshed official Figshare v1 and Zenodo v1/v3 API metadata over verified TLS, including current file inventories, provider checksums and record-level license declarations. The Figshare landing-page description exposes a filename convention and warns of known naming discrepancies. No Figshare or Zenodo payload, article supplement, audio, source code or model weights were downloaded or reproduced; the new publisher search extract was used only to confirm the 2026 article's title, framing and reported result excerpt.

Read our interpretation and limits →

This note cites:
A zebra-finch classifier is a baseline, not a translator

Reviewed the publisher-hosted article's abstract, data/methods, bird-level partition, evaluation, results, limitations, conclusion and data statement on October 24, plus the article-specific version history and AVEX documentation. On October 25, rechecked the publisher-indexed title, framing and result excerpt; the direct MDPI page returned HTTP 429 during this refresh, so no new full-text review is claimed. The official Figshare v1 and Zenodo v1/v3 APIs were refreshed over verified TLS. No archive, code, audio or BirdAVES weights were acquired or run, and no model result was reproduced.

Read our interpretation and limits →

This note cites:
A global bird-song study finds shared acoustic patterns—not a universal bird language

The PubMed abstract reports an analysis of acoustic song architecture across more than 3,000 passerine species and eight recurring motifs associated with species traits and sound-propagation conditions. The authors’ repository README identifies 116,792 song bouts from 3,160 species, describes modulation-power-spectrum features and lists analysis scripts and derived-data files. A recursive GitHub API tree check at pinned commit 1629e665f902cad405ff2f7bf2a0e9dc8c09b47e was not truncated and listed 1,456 tracked files, including 1,280 audio segments totaling 185,758,300 bytes under `data/selected_data_8x8x20_top/`; the tree also lists a 12,796,164-byte `metadata_xeno_passerines.csv` and a 79,543,989-byte `traits_data.csv`. These are provider-reported file sizes and paths only: no file contents or audio were fetched. The tree has no standalone LICENSE file, and the README’s Zenodo data link is the literal placeholder 10.5281/zenodo.XXXXXXX. The GBIF Xeno-canto dataset record describes a metadata subset and a dataset-level CC BY-NC 4.0 declaration. Two unrelated Xeno-canto record pages display different recording licenses; their membership in this study was not checked. No full methods/supplement review, recording-level source/license mapping, data acquisition, code execution, model run or independent reproduction was performed. A follow-up filename screen matched all 1,280 audio paths to a numeric-ID-like segment pattern, yielding 495 distinct candidates. Three path-selected record pages confirmed candidate numbers resolve to Xeno-canto pages; one filename taxon matched the page taxon. Their sample terms were CC BY-NC-ND 4.0 (two) and CC BY-NC-SA 4.0 (one). This validates only the pattern and three examples, not all candidate joins or study membership. The official Xeno-canto API documentation requires a key and warns of strict rate limits; no key or bulk API request was used. No coordinates were extracted or retained.

Read our interpretation and limits →

This note cites:
Synchronized finch recordings help locate calls, not translate them

Reviewed the PubMed bibliographic record and abstract, searchable PMC passages on shared-clock synchronization, multimodal vocal segmentation and channel-miss evaluation, the ETH Research Collection article-rights record, and exact Zenodo v2 descriptions and file inventories for records 13144875 and 20608098. The ETH Zurich dataset record was reviewed for its description and rights label. The complete paper, supplement, annotation guides, CSV contents, audio/video/accelerometer files, code and model weights were not inspected or acquired. No independent avian or ML/statistical reviewer assessed this note.

Read our interpretation and limits →

This note cites:
Human-guided clustering helps label zebra-finch calls, not decode meaning

Reviewed the full 15-page article text, including methods, reported ARI/NMI results, time comparison, split design, discussion, acknowledgments and data statement. Queried exact Zenodo record and version APIs for 18156254, 22657329 and 18100613; read the pinned GitHub README, root requirements and MIT license; and compared the exact WhisperSeg model page with pinned revision metadata. No data archive, code, model weights or supplement was downloaded or executed. Figures were not visually reanalyzed and no independent specialist reviewed this note.

Read our interpretation and limits →

This note cites:
Cross-species pitch benchmark helps estimate vocal pitch, not meaning

Reviewed the full 2025 PMC article, Dryad landing page and official dataset/version/file-manifest APIs, Dryad reuse guidance, and the pinned code tree. For source lineage, reviewed the 2025 paper's monk-parakeet table/references, the 2021 Animal Behaviour publisher abstract/data-accessibility statement and Figshare API manifest, plus the 2023 PLOS article record, Figshare v2 API and streamed README. The Dryad API snapshot is saved in research/dryad_f0_manifest_2026-10-24.json. No Dryad archive, audio, selection-table payload, code or weights were acquired or run; no Figshare rows or KML contents were inspected.

Read our interpretation and limits →

This note cites:
A large insect sound dataset can test species recognition—not what a call means

The paper abstract reports 26,298 recordings (226.6 hours) from 459 Orthoptera and Cicadidae species. The Zenodo v1.1 record lists an 83.7 GB package, including a 6.9 MB annotation CSV, and describes source recordings as CC licenses 4.0 or CC0. The record states a 60/20/20 split balanced by file number and duration. The paper describes broad recording sample rates, from 8 to 500 kHz. Approximately one third of files use sample rates above 44 kHz according to the paper. The full paper methods, split membership, original annotation rows, source-level license manifest, location values and code were not inspected.

Read our interpretation and limits →

This note cites:
Which companion animal is ready for a first study?

Compared the existing full-paper review of dog-bark classification; CatMeows record and rights metadata; full-text review and deposit metadata for two companion-parrot surveys; primary methods and official deposit inventory for budgerigar vocal/sociality data; and the publisher abstract for a 2026 edge-monitoring study. Supporting survey workbooks and third-party media were not acquired or inspected. This is a library-bounded comparison, not a systematic review or independent specialist assessment.

Read our interpretation and limits →

This note cites:
Honeybee dance records connect movement with followers and food sites

Targeted review of the PNAS Nexus abstract, communication background, colony and camera setup, annotation and hourly sample grouping, follower/feeder results, data statement and article license; on 2026-10-25, rechecked the complete Zenodo v1.0 API file manifest and dataset license. The 11 listed sizes sum to 4,870,358,178 bytes; the named follower CSV is 9,171,824 bytes. No data, video, code or model was downloaded, locally hashed, inspected or executed.

Read our interpretation and limits →

This note cites:
MooLogue separates cattle-call learning from live analysis

Reviewed selected abstract, app design and release, curated-corpus description, dataset-access statement and article license in an author-uploaded article copy. Read the README and README_softwarex.md at pinned GitHub commit 8023dbdfed8587d07eeafa54accb68408ffaefbe. The Zenodo API metadata request returned HTTP 406, so its manifest was not independently inspected. No app was installed, no repository archive or dataset payload was acquired, and no code or model was run.

Read our interpretation and limits →

This note cites:
Nocturnal bird migration data localizes calls, not communication meaning

Reviewed the Scientific Data article's abstract, methods, data records, technical validation, model-evaluation description, data/code availability and article-rights statement; the current Zenodo v4 landing page and file list; the linked GitHub repository root and README; and the Zenodo software-record page. No dataset file, metadata CSV, audio, code archive or model weight was downloaded, opened, installed or executed. No result was independently reproduced.

Read our interpretation and limits →

This note cites:
BirdNET Live keeps bird identification on the phone

Reviewed official Getting Started and Privacy pages (policy updated 2026-08-06) plus the GitHub v1.2.1 release at commit 4aee47cd824873ffc9b3ca3462bf2415151a56d1, including README, LICENSE and MODEL_LICENSE. Documentation describes app flows and developer data-handling claims; these were not independently runtime-tested. No app, source archive, model, audio or training data was acquired or run.

Read our interpretation and limits →

This note cites:
Island parakeet contact calls vary across locations

Reviewed the publisher-indexed abstract, recording approach, Table 1 location statement, frequency/SPCC-PCO methods, duration results and selected discussion passages; this is targeted review, not a complete article audit. The official Dryad API v2 record identifies dataset 77790, version 4 / version ID 147704, dates, dataset-level CC0-1.0 and eight file records with declared SHA-256 digests. The direct publisher fetch returned HTTP 403. The official README API download returned HTTP 401 and its record-page download link returned HTTP 403. No article PDF, README, spreadsheet, audio, supplement or code was acquired; upstream hashes were not locally checked.

Read our interpretation and limits →

This note cites:

A foundation for shared research.

These connections are citation records, not statements that every source agrees. Each record has a content fingerprint; a changed source record produces a different version.

The private prototype also connects frozen study references, sessions, observations and releases. It excludes withdrawn observations from its graph. Media storage and model-run records are the next schema extensions, not live public services.

Advanced · machine-readable catalog

Download the public graph · JSON

Schema: talk2nature.knowledge.v1. Original catalog metadata only; this file is not a training dataset or permission to reuse third-party works. Citation targets reflect the catalog at build time, not a verified historical version of the original webpage.

Explore the scientific direction →