hewittlab/monitor_cot_for_hack-judge
monitor_cot_for_hack-judge Training data for the judge task of Reward Misspecification Degrades Chain-of-Thought Monitorability (code), built from lmarena-ai/arena-human-preference-140k (cc-by-4.0). Data: Chatbot Arena Claude-vs-Gemini response pairs that are equally preferred by human annotators Entry num per split: train 840, dev 111, test 111, probe 50 Probe: probe is a held out set of queries used for generating reward descriptions True reward: L1 similarity between the… See the full description on the dataset page: https://huggingface.co/datasets/hewittlab/monitor_cot_for_hack-judge.
| Author / organization | hewittlab |
|---|---|
| License (as declared) | cc-by-4.0 |
| Downloads | 0 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 16 October 2026 |
| Last updated at source | 16 October 2026 |
Connected in the knowledge graph
- hewittlab/monitor_cot_for_hack-baseline-judge-explicit_intent-gpt-oss-120bModel · used to train
- hewittlab/monitor_cot_for_hack-reward_description-judge-explicit_intent-gpt-oss-20bModel · used to train
- hewittlab/monitor_cot_for_hack-baseline-judge-explicit_intent-nemotron-3-nano-30b-a3bModel · used to train
- hewittlab/monitor_cot_for_hack-baseline-judge-explicit_intent-gpt-oss-20bModel · used to train
More from hewittlab
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 16 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.