Claude 3 Opus strategically fakes alignment by complying with harmful requests only during simulated training to preserve its preference for refusing them afterward.
mega hub Mixed citations
Proximal Policy Optimization Algorithms
Mixed citation behavior. Most common role is background (52%).
1867
Pith papers citing it
Background
52% of classified citations
abstract
We propose a new family of policy gradient methods for reinforcement learning, which alternate between sampling data through interaction with the environment, and optimizing a "surrogate" objective function using stochastic gradient ascent. Whereas standard policy gradient methods perform one gradient update per data sample, we propose a novel objective function that enables multiple epochs of minibatch updates. The new methods, which we call proximal policy optimization (PPO), have some of the benefits of trust region policy optimization (TRPO), but they are much simpler to implement, more general, and have better sample complexity (empirically). Our experiments test PPO on a collection of benchmark tasks, including simulated robotic locomotion and Atari game playing, and we show that PPO outperforms other online policy gradient methods, and overall strikes a favorable balance between sample complexity, simplicity, and wall-time.
hub tools
citation-role summary
background 155
method 113
baseline 15
dataset 4
citation-polarity summary
claims ledger
- abstract We propose a new family of policy gradient methods for reinforcement learning, which alternate between sampling data through interaction with the environment, and optimizing a "surrogate" objective function using stochastic gradient ascent. Whereas standard policy gradient methods perform one gradient update per data sample, we propose a novel objective function that enables multiple epochs of minibatch updates. The new methods, which we call proximal policy optimization (PPO), have some of the benefits of trust region policy optimization (TRPO), but they are much simpler to implement, more ge
authors
mega hub controls
export citing contexts JSON
export graph JSON
export full bundle JSON
open full Pith review
annotated reader queued
Recognition alignment
counterfactual ablation
co-cited works
representative citing papers
GRPO's group-mean baseline assigns identical advantages to all tokens under output-only rewards, inducing gradient sparsity and an intrinsic rank-2 structure proven from the zero-sum constraint and confirmed by SVD on Nemotron-4B gradients.
Negative narrative immersion causes 12-31% drops in LLM moral accuracy and produces structured shifts that appear in downstream applications.
RL agent for online LHC trigger threshold tuning improves in-tolerance intervals by 28-56% on Monte Carlo and real CMS data without fine-tuning.
DeFAb is a large-scale, formally verifiable benchmark for defeasible abduction derived from 18 knowledge bases, demonstrating that frontier LLMs achieve 7.8-65% accuracy versus 100% for a rule-based solver with polynomial-time checks.
Machine learning discovers a tube-seeding strategy for IBP reduction of Feynman integrals that scales linearly with numerator power, demonstrated on rank-20 2-loop 5-point integrals.
A reward-free representation learning pipeline for offline PbRL achieves better preference efficiency than standard two-stage baselines by connecting RFRL concepts to preference data.
Dynamic isotropy, quantifying uniform center-of-mass acceleration capability, improves robot performance and enables omnidirectional locomotion, terrain traversal, and failure resilience in a spherical robot design.
AtomComposer uses online RL with multi-composition training to discover up to 10x more valid 3D isomers on unseen chemical formulas than single-composition baselines.
Agent-BRACE improves LLM agent performance on long-horizon partially observable tasks by 5.3-14.5% through a decoupled belief state of verbalized atomic claims with certainty labels that keeps context length constant.
SimWorld Studio deploys an evolving coding agent to create adaptive 3D environments that co-evolve with embodied learners, delivering 18-point success-rate gains over fixed environments in navigation benchmarks.
ReLibra uses pre-known token-to-expert routing from RL rollouts to perform inter-batch expert reordering and intra-batch replication, delivering up to 1.6x higher throughput than Megatron-LM and 1.2x over oracle-equipped EPLB while staying within 6-10% of an ideal balanced baseline.
Weak-to-strong generalization is nearly inevitable in linear logistic regression for most student-teacher pairs without any model capacity mismatch.
Observation and action delays are formally equivalent in cooperative Dec-POMDPs, yielding identical optimal solutions and enabling zero-shot transfer, though learning dynamics differ due to credit assignment and operational constraints.
A language-game framework enables dialogue with dynamical systems such as GRNs by treating their frozen dynamics as an RL policy core, using an LM to route prompts so the system responds through its own behavior without parameter changes.
RefereeBench shows that even the strongest video MLLMs reach only around 60% accuracy on multi-sport refereeing tasks and struggle with rule application and temporal grounding.
OP-GRPO is the first off-policy GRPO method for flow-matching models that reuses trajectories via replay buffer and importance sampling corrections, matching on-policy performance with 34.2% of the training steps.
User-turn generation reveals that LLMs' interaction awareness is largely decoupled from task accuracy, remaining near zero in deterministic settings even as accuracy scales to 96.8% on GSM8K.
Derives an exact unbiased policy gradient for RL post-training of diffusion LLMs via entropy-guided step selection and one-step denoising rewards, achieving state-of-the-art results on coding and logical reasoning benchmarks.
A certified gradient-based method for contact-rich manipulation that quantifies smoothing-induced errors via set-valued discrepancies and incorporates them into analytical reachable sets for robust affine feedback policies.
First in-orbit demonstration of a DRL-trained AI satellite attitude controller that performs robust inertial pointing after sim-to-real transfer.
Develops and tests the first effective safeguard for analytic gradient-based provably safe RL, showing safe training on three control tasks without performance loss.
Flow-GRPO is the first online RL method for flow matching models, raising GenEval accuracy from 63% to 95% and text-rendering accuracy from 59% to 92% with little reward hacking.
SWE-Gym supplies 2438 executable real-world Python tasks to train SWE agents and verifiers, yielding up to 19% gains and new open-weight SOTA of 32% on SWE-Bench Verified.
citing papers explorer
-
Agent-BRACE: Decoupling Beliefs from Actions in Long-Horizon Tasks via Verbalized State Uncertainty
Agent-BRACE improves LLM agent performance on long-horizon partially observable tasks by 5.3-14.5% through a decoupled belief state of verbalized atomic claims with certainty labels that keeps context length constant.
-
ORPO: Monolithic Preference Optimization without Reference Model
ORPO performs preference alignment during supervised fine-tuning via a monolithic odds ratio penalty, allowing 7B models to outperform larger state-of-the-art models on alignment benchmarks.
-
Zone of Proximal Policy Optimization: Teacher in Prompts, Not Gradients
ZPPO improves distillation to small vision-language models by using binary and negative candidate prompts plus a replay buffer for hard questions, outperforming standard distillation and GRPO on a 31-benchmark suite with largest gains at the 0.8B scale.
-
Hey Chat, Can You Teach Me? Structuring Socratic Dialogue for Human Learning in the Wild
A PPO policy for deciding topic order and duration on a prerequisite knowledge graph, paired with an LLM for Socratic dialogue, improves student mastery rates and reduces turns compared to baselines and scaled models across held-out topics.
-
ACE-SQL: Adaptive Co-Optimization via Empirical Credit Assignment for Text-to-SQL
ACE-SQL jointly optimizes schema linking and SQL generation via RL with empirical credit assignment from execution-correct rollouts, achieving 65.3% greedy execution accuracy on BIRD Dev using 0.93k output tokens.
-
Generative Floor Plan Design with LLMs via Reinforcement Learning with Verifiable Rewards
Fine-tuned LLMs trained with reinforcement learning using verifiable rewards produce floor plans that satisfy connectivity and numerical constraints, outperforming prior methods with at least 94% relative improvement in compatibility.
-
Simulating Students or Sycophantic Problem Solving? On Misconception Faithfulness of LLM Simulators
LLM simulators exhibit near-zero selective response to targeted misconception feedback and behave sycophantically, but SFT and SFS-aligned RL improve this property.
-
TokenRatio: Principled Token-Level Preference Optimization via Ratio Matching
Introduces TBPO, which derives a Bregman-divergence density-ratio matching objective for token-level preference optimization that generalizes DPO while preserving the induced optimal policy.
-
Learning Agentic Policy from Action Guidance
ActGuide-RL uses human action data as plan-style guidance in mixed-policy RL to overcome exploration barriers in LLM agents, matching SFT+RL performance on search benchmarks without cold-start training.
-
DeepRefine: Agent-Compiled Knowledge Refinement via Reinforcement Learning
DeepRefine refines agent-compiled knowledge bases via multi-turn abductive diagnosis and RL training with a GBD reward, yielding consistent downstream task gains.
-
Relative Score Policy Optimization for Diffusion Language Models
RSPO interprets reward advantages as targets for relative log-ratios in dLLMs, calibrating noisy estimates to stabilize RLVR training and achieve strong gains on planning tasks with competitive math reasoning performance.
-
Guidance Is Not a Hyperparameter: Learning Dynamic Control in Diffusion Language Models
Adaptive guidance trajectories learned via PPO outperform fixed-scale CFG on controllability-quality balance in three controlled NLP generation tasks with discrete diffusion models.
-
Beyond Reasoning: Reinforcement Learning Unlocks Parametric Knowledge in LLMs
RL on binary rewards boosts LLM factual recall by ~27% relative across models by redistributing probability mass to latent correct answers rather than acquiring new knowledge.
-
Beyond Negative Rollouts: Positive-Only Policy Optimization with Implicit Negative Gradients
POPO uses bounded importance sampling on positive rollouts and a siamese policy network to achieve implicit negative gradients and stable optimization, matching or exceeding GRPO on math benchmarks such as 36.67% on AIME 2025.
-
Enhanced LLM Reasoning by Optimizing Reward Functions with Search-Driven Reinforcement Learning
Iterative search over reward functions with ranked feedback in GRPO training improves LLM math reasoning, achieving F1 of 0.795 on GSM8K versus 0.609 for baseline.
-
A Multi-View Media Profiling Suite: Resources, Evaluation, and Analysis
Presents MBFC-2025 dataset and multi-view embeddings with fusion methods for media bias and factuality, reporting SOTA results on ACL-2020 and new benchmarks on MBFC-2025.
-
RSAT: Structured Attribution Makes Small Language Models Faithful Table Reasoners
RSAT uses SFT on verified traces followed by GRPO with NLI faithfulness rewards to make 1-8B models produce verifiable table reasoning with cell citations, raising faithfulness 3.7x to 0.826.
-
TLPO: Token-Level Policy Optimization for Mitigating Language Confusion in Large Language Models
TLPO mitigates language confusion in LLMs via token-level policy updates that outperform sequence-level methods while preserving general capabilities.
-
EXPO-SQL: Execution-based Clause-level Policy Optimization for Text-to-SQL
EXPO-SQL improves Text-to-SQL by using clause-level rewards derived from execution error messages and incremental clause execution instead of uniform query-level rewards.
-
ReflectMT: Internalizing Reflection for Efficient and High-Quality Machine Translation
ReflectMT internalizes reflection via two-stage RL to enable direct high-quality machine translation that outperforms explicit reasoning models like DeepSeek-R1 on WMT24 while using 94% fewer tokens.
-
Reverse Constitutional AI: A Framework for Controllable Toxic Data Generation via Probability-Clamped RLAIF
R-CAI inverts constitutional AI to automatically generate diverse toxic data for LLM red teaming, with probability clamping improving output coherence by 15% while preserving adversarial strength.
-
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.
-
Foresight Optimization for Strategic Reasoning in Large Language Models
FoPO trains LLMs for strategic reasoning by combining self-interest with opponent modeling in policy optimization, yielding gains on two new datasets and better out-of-domain generalization than standard baselines.
-
Teaching LLMs Human-Like Editing of Inappropriate Argumentation via Reinforcement Learning
Reinforcement learning with a multi-part reward teaches LLMs to output independent, meaning-preserving sentence edits that raise argument appropriateness close to full rewriting.
-
Demystifying OPD: Length Inflation and Stabilization Strategies for Large Language Models
OPD for LLMs suffers length inflation and repetition collapse; StableOPD uses reference divergence and rollout mixing to prevent it and improve math reasoning performance by 7.2% on average.
-
Right at My Level: A Unified Multilingual Framework for Proficiency-Aware Text Simplification
Re-RIGHT trains a 4B policy model with vocabulary coverage, semantic preservation, and coherence rewards to perform proficiency-aware lexical simplification in four languages without parallel corpora.
-
Many Preferences, Few Policies: Towards Scalable Language Model Personalization
PALM produces a small portfolio of LLMs that contains a near-optimal model for any user preference weight vector, with theoretical bounds on portfolio size and approximation quality.
-
Temp-R1: A Unified Autonomous Agent for Complex Temporal KGQA via Reverse Curriculum Reinforcement Learning
Temp-R1 uses reverse curriculum reinforcement learning to train an autonomous agent that achieves state-of-the-art results on temporal KGQA benchmarks by developing sophisticated reasoning on hard questions first.
-
PIAST: Rapid Prompting with In-context Augmentation for Scarce Training data
PIAST iteratively optimizes few-shot examples in prompts via Monte Carlo Shapley value estimation, outperforming prior automatic prompting methods and setting new SOTA on classification, simplification, and GSM8K with modest compute.
-
MemSearcher: Training LLMs to Reason, Search and Manage Memory via End-to-End Reinforcement Learning
MemSearcher trains LLMs to manage compact memory in multi-turn searches via multi-context GRPO for end-to-end RL, outperforming ReAct-style baselines with stable token counts.
-
HiPRAG: Hierarchical Process Rewards for Efficient Agentic Retrieval Augmented Generation
HiPRAG adds hierarchical process rewards to RL training for agentic RAG, reducing over-search to 2.3% and achieving 65.4-67.2% accuracy on seven QA benchmarks across 3B and 7B models.
-
Self-Rewarding Language Models
Iterative self-rewarding via LLM-as-Judge in DPO training on Llama 2 70B improves instruction following and self-evaluation, outperforming GPT-4 on AlpacaEval 2.0.
-
Catastrophic Jailbreak of Open-source LLMs via Exploiting Generation
Varying decoding strategies such as temperature and sampling methods jailbreaks safety alignments in open-source LLMs, raising misalignment from 0% to over 95% at 30x lower cost than prior attacks.
-
Learning to summarize from human feedback
Reinforcement learning on a reward model trained from human summary comparisons produces summaries humans prefer over supervised fine-tuning or human references on TL;DR and transfers to CNN/DM.
-
Fine-Tuning Language Models from Human Preferences
Language models fine-tuned via RL on 5k-60k human preference comparisons produce stylistically better text continuations and human-preferred summaries that sometimes copy input sentences.
-
Training Therapeutic Judges and Multi-Agent Systems for Human-Aligned Mental Health Support
TheraJudge, trained via preference optimization on human annotations, reaches high clinician agreement (ICC 0.87-0.95) and, when used by TheraAgent, raises human-rated therapeutic quality by 0.43 points on a 5-point scale with 94% recovery of low-quality responses.
-
Before Thinking, Learn to Decide: Proactive Routing for Efficient Visual Reasoning
PRP introduces proactive routing via Draft Rating Learning and Joint Rating Learning to route queries early between draft and target models for efficient multimodal reasoning.
-
ToxiREX: A Dataset on Toxic REasoning in ConteXt
ToxiREX is a new dataset of 128k Reddit comments in six languages with hierarchical annotations for implicit toxicity in conversational context based on an existing reasoning schema.
-
The Wrong Kind of Right: Quantifying and Localizing Misfired Alignment in LLMs
LLMs exhibit misfired alignment on stereotype questions at 4.7-18.9% rates on the new VETO benchmark of 2,032 contrastive pairs, unlike humans at 0%, due to overgeneralized safety cues after instruction tuning.
-
Learning to Reason by Analogy via Retrieval-Augmented Reinforcement Fine-Tuning
RA-RFT trains a retriever to rank contexts by expected reasoning benefit and uses the retrieved analogies inside reinforcement fine-tuning, yielding 7.1 and 2.8 point gains on AIME 2025 over GRPO for two Qwen3 models.
-
NaturalFlow: Reducing Disruptive Pauses for Natural Speech Flow in Simultaneous Speech-to-Speech Translation
A fluency-aware optimization framework is introduced to minimize inter-chunk silences in simultaneous speech-to-speech translation by leveraging model-internal signals including linguistic diversity and temporal variability.
-
Beyond Fully Random Masking: Attention-Guided Denoising and Optimization for Diffusion Language Models
AGDO improves dLLM reasoning performance by determining denoising order and emphasizing tokens based on attention-derived dependencies rather than random masking.
-
On The Effectiveness-Fluency Trade-Off In LLM Conditioning: A Systematic Study
Systematic experiments reveal that activation steering trades fluency for concept control, is less effective on instruction-tuned models, and that prompting/SFT excel at injection but not removal, with textual metrics correlating to LLM judges.
-
Multi-Faceted Interactivity Alignment in Full-Duplex Speech Models
A multi-axis RL alignment technique improves pause handling, turn-taking, backchanneling, and interruption response in full-duplex spoken dialogue models by optimizing axis-specific rewards derived from human audio segments.
-
Learning to Attack and Defend: Adaptive Red Teaming of Language Models via GRPO
AdvGRPO stabilizes GRPO for joint attacker-defender optimization via multi-channel rewards and curriculum training, yielding effective transferable attacks and stronger co-trained defenders on safety benchmarks.
-
Multilingual Sentiment Aware Text Summarization A Reinforcement Learning Approach for Consistency Maintenance
RLHF causes consistent sentiment drift to neutral in multilingual summarization primarily via KL regularization, which a sentiment-aware KL modification can mitigate.
-
Back on Track: Aligning Rewards and States for Reasoning in Diffusion Large Language Models
PAPO improves reasoning performance in diffusion LLMs by converting sparse terminal rewards into dense step-wise credit and replaying real high-uncertainty trajectories, reporting gains up to 42.2% on Countdown.
-
CATPO: Critique-Augmented Tree Policy Optimization
CATPO introduces an informativeness score F(T) and critique-guided healing for failed trees to improve efficiency and performance in tree-based RLVR, reaching 37.5% macro accuracy on math benchmarks.
-
Agentopia: Long-Term Life Simulation and Learning in Agent Societies
Agentopia runs decade-scale multi-agent LLM simulations to study emergent social behaviors and trains models with life-reward rejection sampling, yielding +15.6% gains on role-playing benchmarks.
-
MADE: Beyond Scoring via a Multilingual Agentic Diagnosing Engine for Fine-Grained Evaluation Insights
MADE is a new multilingual agentic diagnosing engine that produces higher-quality diagnostic reports (47% better than baseline) on a large-scale evaluation substrate covering 33 model families and 26 languages.