McGill-NLP/A3-Synth
A3-Synth πΎ Code π Paper π Website π€ Dataset π€ Models π¦ PyPI Structured Distillation of Web Agent Capabilities Enables Generalization Xing Han LΓΉ, Siva Reddy A3-Synth is a synthetic training dataset for web agents, generated using the Agent-as-Annotators (A3) framework. It contains ~16k SFT training examples produced by Gemini 3 Pro acting as the Annotator across 3,000 tasks on 6 WebArena environments. Dataset Structure A3-Synth/β¦ See the full description on the dataset page: https://huggingface.co/datasets/McGill-NLP/A3-Synth.
| Author / organization | McGill-NLP |
|---|---|
| Downloads | 4.9K (as reported by Hugging Face) |
| Likes | 1 |
| Created | 6 April 2026 |
| Last updated at source | 10 April 2026 |
Popularity, last 3 days
4,946 now (-55 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Connected in the knowledge graph
Recent changes
- 16 Oct 2026 Description: A3-Synth πΎ Code π Paper π Website π€ Dataset π€ Models π¦ PyPI Structured Distillation of Web Agent Capabilities Enables Generalization Xing Han LΓΉ, Siva Reddy A3-Synth is a synthetic training dataset for web agents, generated using the β A3-Synth πΎ Code π Paper π Website π€ Dataset π€ Models π¦ PyPI Structured Distillation of Web Agent Capabilities Enables Generalization Xing Han LΓΉ, Siva Reddy A3-Synth is a synthetic training dataset for web agents, generated using the
Changes detected by Silytics between refreshes of the source.
Related papers
More from McGill-NLP
- McGill-NLP/agent-reward-benchDataset
- McGill-NLP/WebLINX-fullDataset
- McGill-NLP/SpeechJBBDataset
- McGill-NLP/weblinx-browsergymDataset
- McGill-NLP/stereosetDataset
- McGill-NLP/WebLINXDataset
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.