The cumulative token IS ratio gives unbiased prefix correction and lower variance than full-sequence ratios for token-level gradients in LLM policy optimization, enabling CTPO to outperform GRPO and GSPO baselines on mathematical reasoning tasks.
hub
Rlhf workflow: From reward modeling to online rlhf.arXiv preprint arXiv:2405.07863
11 Pith papers cite this work. Polarity classification is still indexing.
11
Pith papers citing it
hub tools
citation-role summary
background 1
citation-polarity summary
verdicts
UNVERDICTED 11roles
background 1polarities
background 1representative citing papers
Freshness-Aware PER augments prioritized experience replay with exponential age decay based on effective sample size to enable successful reuse of trajectories in LLM and VLM reinforcement learning, outperforming on-policy baselines on agentic tasks.
SelfGrader detects LLM jailbreaks by interpreting logit distributions on numerical tokens with a dual maliciousness-benignness score, cutting attack success rates up to 22.66% while using up to 173x less memory and 26x less latency.
MORA breaks the safety-helpfulness ceiling in LLMs by pre-sampling single-reward prompts and rewriting them to incorporate multi-dimensional intents, delivering 5-12.4% gains in sequential alignment and 4.6% overall improvement in simultaneous alignment.
SelectiveRM applies optimal transport with a joint consistency discrepancy and partial mass relaxation to produce reward models that optimize a tighter upper bound on clean risk while autonomously dropping noisy preference samples.
PERSA combines RLHF with selective parameter-efficient updates to top transformer layers, raising style alignment scores from 35% to 96% on code feedback benchmarks while holding correctness near 100%.
SceneCritic is a symbolic, ontology-grounded evaluator for floor-plan layouts that identifies specific semantic, orientation, and geometric violations and aligns better with human judgments than VLM-based scorers.
A user-diversity condition is necessary and sufficient for personalized alignment to achieve O(1) online regret and log(1/epsilon) offline sample complexity.
The paper introduces the Proxy Compression Hypothesis as a unifying framework explaining reward hacking in RLHF as an emergent result of compressing high-dimensional human objectives into proxy reward signals under optimization pressure.
The paper unifies perspectives on Long CoT in reasoning LLMs by introducing a taxonomy, detailing characteristics of deep reasoning and reflection, and discussing emergence phenomena and future directions.
Curr-RLCER applies curriculum reinforcement learning with coherence-driven rewards to align generated explanations with predicted ratings in explainable recommendation systems.
citing papers explorer
-
Rethinking Importance Sampling in LLM Policy Optimization: A Cumulative Token Perspective
The cumulative token IS ratio gives unbiased prefix correction and lower variance than full-sequence ratios for token-level gradients in LLM policy optimization, enabling CTPO to outperform GRPO and GSPO baselines on mathematical reasoning tasks.
-
Freshness-Aware Prioritized Experience Replay for LLM/VLM Reinforcement Learning
Freshness-Aware PER augments prioritized experience replay with exponential age decay based on effective sample size to enable successful reuse of trajectories in LLM and VLM reinforcement learning, outperforming on-policy baselines on agentic tasks.
-
SelfGrader: Stable Jailbreak Detection for Large Language Models using Token-Level Logits
SelfGrader detects LLM jailbreaks by interpreting logit distributions on numerical tokens with a dual maliciousness-benignness score, cutting attack success rates up to 22.66% while using up to 173x less memory and 26x less latency.
-
Explaining and Breaking the Safety-Helpfulness Ceiling via Preference Dimensional Expansion
MORA breaks the safety-helpfulness ceiling in LLMs by pre-sampling single-reward prompts and rewriting them to incorporate multi-dimensional intents, delivering 5-12.4% gains in sequential alignment and 4.6% overall improvement in simultaneous alignment.
-
Optimal Transport for LLM Reward Modeling from Noisy Preference
SelectiveRM applies optimal transport with a joint consistency discrepancy and partial mass relaxation to produce reward models that optimize a tighter upper bound on clean risk while autonomously dropping noisy preference samples.
-
PERSA: Reinforcement Learning for Professor-Style Personalized Feedback with LLMs
PERSA combines RLHF with selective parameter-efficient updates to top transformer layers, raising style alignment scores from 35% to 96% on code feedback benchmarks while holding correctness near 100%.
-
SceneCritic: A Symbolic Evaluator for 3D Indoor Scene Synthesis
SceneCritic is a symbolic, ontology-grounded evaluator for floor-plan layouts that identifies specific semantic, orientation, and geometric violations and aligns better with human judgments than VLM-based scorers.
-
Personalized Alignment Revisited: The Necessity and Sufficiency of User Diversity
A user-diversity condition is necessary and sufficient for personalized alignment to achieve O(1) online regret and log(1/epsilon) offline sample complexity.
-
Reward Hacking in the Era of Large Models: Mechanisms, Emergent Misalignment, Challenges
The paper introduces the Proxy Compression Hypothesis as a unifying framework explaining reward hacking in RLHF as an emergent result of compressing high-dimensional human objectives into proxy reward signals under optimization pressure.
-
Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Models
The paper unifies perspectives on Long CoT in reasoning LLMs by introducing a taxonomy, detailing characteristics of deep reasoning and reflection, and discussing emergence phenomena and future directions.
-
Curr-RLCER:Curriculum Reinforcement Learning For Coherence Explainable Recommendation
Curr-RLCER applies curriculum reinforcement learning with coherence-driven rewards to align generated explanations with predicted ratings in explainable recommendation systems.