AI Paper Atlas

Research map / Large language models / Alignment and preferences

Long-form RewardBench: Evaluating Reward Models for Long-form Generation

Hui Huang, Yancheng He, Wei Liu, Muyun Yang, Jiaheng Liu, Kehai Chen, Bing Xu, Conghui Zhu, Hailong Cao, Tiejun Zhao

AAAI 2026 · Large language models › Alignment and preferences · approach: Reward model design and preference models · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
evaluate generative reward models in Selection mode instead of Scoring mode (ablation)generative reward models · Long-form RewardBenchoverall accuracy: gemini-2.5-flash 40.8 -> 73.6; gpt-4o 41.1 -> 60.9Improved
use sequence classifiers instead of generative models for overall long-form reward modelingreward models · Long-form RewardBenchoverall accuracy: top classifier 74.9 vs best generative 74.6Improved
use generative models (with chain-of-thought) instead of classifiers on reasoning subsetreward models · Long-form RewardBench Reasoningpairwise accuracy: classifiers about 70% vs generative over 90% (gemini-2.5-flash 90.2)Improved
use fine-tuned generative judges (RISE, Selene, Skywork-Critic) for best-of-n and pointwise scoringfine-tuned generative judges · Long-form RewardBench BoN subsetsvalid output rate: invalid up to 73.9% (Skywork-Critic Scoring)Hurt
increase response length from 1K to 8K tokens in needle test (ablation)generative RM (GPT-4o) · Long-form Needle-in-a-Haystackerror detection accuracyHurt
place the error in the middle of the response (ablation)generative RM (GPT-4o) · Long-form Needle-in-a-Haystackerror detection accuracyHurt
increase response length (ablation)sequence classifier (Skywork-Llama-3.1-8B-v0.2) · Long-form Needle-in-a-Haystackerror detection accuracyNo clear effect
place the error near the end of the response (ablation)sequence classifier · Long-form Needle-in-a-Haystackerror detection accuracyImproved
train on longer preference data (ablation)classifier reward models · Long-form RewardBench, 30k synthetic pairs split by lengthaccuracyNo clear effect
train on longer preference data (ablation)generative reward models · Long-form RewardBench, 30k synthetic pairs split by lengthaccuracyImproved
use Qwen base instead of Llama base for generative reward model (ablation)generative reward models · Long-form RewardBenchaccuracyImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy