Dataset · allenai
RewardBench benchmark for evaluating reward models
allenai/reward-bench
Dataset for testing reward models on chats, complex dialogues, and preferences; English text only.
What's inside
RewardBench is an evaluation dataset for assessing reward model capabilities. The card lists Chat categories with simple chat subsets (alpacaeval-easy, alpacaeval-length, alpacaeval-hard, mt-bench-easy, mt-bench-medium) and Chat Hard with complex ones (mt-bench-hard, llmbar-natural, llmbar-adver-neighbor, llmbar-adver-GPTInst, llmbar-adver-GPTOut). The language is English; the format is not detailed in the card.
What you can do with it
- Evaluating and comparing reward models for RLHF
- Choosing a reward model prior to PPO/DPO training
- Testing model robustness on complex and adversarial dialogues
- Tasks
- Question answering
- Languages
- English
- Size
- 1K–10K rows
- Downloads this month
- 19.8K
- Likes
- 113
Models trained on it
- samratduttaofficial/WaterSheepZero shot classification 87
- haizelabs/j1-nano-0.6B 70
- ITLL/WaterSheepZero shot classification 32
- mario-rc/multi-domain-rm-mistral-7b-itText classification 30
- mario-rc/multi-domain-rm-skywork-llama-3.1-8b-itText classification 29
- haizelabs/j1-micro-1.7B 26
- mario-rc/multi-domain-rm-skywork-qwen-3-8b-itText classification 24
- gyung/Qwev-9B-RLCDText classification 22
How to load it
from datasets import load_dataset
ds = load_dataset("allenai/reward-bench")