AR-OPD disentangles privileged supervision via anchored residual guidance to reduce hindsight leakage in on-policy distillation, reporting gains of 2.3 points over full privileged OPD and 7.9 over SFT on reasoning tasks.
hub
Pope: Learning to reason on hard problems via privileged on-policy exploration
16 Pith papers cite this work. Polarity classification is still indexing.
16
Pith papers citing it
hub tools
citation-role summary
background 3
citation-polarity summary
years
2026 16roles
background 3polarities
background 3representative citing papers
DistIL applies distributional DAgger with forward cross-entropy to achieve monotonic policy improvement and better Pass@N from rich feedback in RL for reasoning tasks.
VPD frames language feedback learning as variational EM so the teacher policy refines itself via trust-region updates on outcomes while the student learns dense token distributions on its own rollouts, outperforming fixed-teacher baselines on reasoning and code tasks.
GUI-SD introduces on-policy self-distillation with visually enriched privileged context and entropy-guided weighting, outperforming GRPO and naive OPSD on six GUI grounding benchmarks while improving training efficiency.
FADE is a self-adapting advantage for policy-gradient RL that reads training dynamics to balance positive/negative gradient mass and difficulty focus, yielding faster peak performance and better accuracy-diversity trade-offs than static baselines on LLM reasoning benchmarks.
REVES augments LLM post-training by decoupling revision and verification signals from successful multi-step trajectories, reporting +6.5 point gains on LiveCodeBench over RL baselines.
AXPO addresses the Thinking-Acting Gap in agentic RL training by targeted resampling of tool calls in all-wrong subgroups, delivering +1.8pp gains over GRPO on nine multimodal benchmarks with an 8B model beating a 32B baseline on Pass@4.
NudgeRL conditions RLVR rollouts on strategy-level contexts to drive diverse trajectories and applies an inter/intra-context reward decomposition plus distillation objective, outperforming GRPO and oracle baselines on math benchmarks.
Prefix Sampling replays self-generated trajectory prefixes to control rollout pass rates near 50% in binary-reward RL, delivering wall-clock speedups and modest performance gains on SWE-bench Verified and AIME tasks.
Uni-OPD unifies on-policy distillation across LLMs and MLLMs with dual-perspective strategies that promote student exploration and enforce order-consistent teacher supervision based on outcome rewards.
SGS adds self-guidance to LLM self-play for Lean4 theorem proving, surpassing RL baselines and enabling a 7B model to outperform a 671B model after 200 rounds.
AutoOR uses synthetic data generation and RL post-training with solver feedback to enable 8B LLMs to autoformalize linear, mixed-integer, and non-linear OR problems, matching larger models on benchmarks.
PolicyLong shifts long-context data synthesis to an on-policy loop that re-screens contexts using the evolving model's entropy landscape, producing a self-curriculum that outperforms static offline baselines with larger gains at longer lengths.
Fine-tuned LLMs trained on social science publication records outperform experts and frontier models at judging which research pitches deserve attention.
SMEPO applies fine-grained semantic masking to expert guidance in RLVR, turning hard problems into fill-in-the-blank tasks while preserving structure, yielding up to 3.2 point accuracy gains and 4.2x faster training.
DenoiseRL optimizes recovery from noisy prefixes in weak-model reasoning failures to improve performance and self-correction on math and general reasoning benchmarks without external supervision.
citing papers explorer
-
Beyond Absolute Imitation: Anchored Residual Guidance for Privileged On-Policy Distillation
AR-OPD disentangles privileged supervision via anchored residual guidance to reduce hindsight leakage in on-policy distillation, reporting gains of 2.3 points over full privileged OPD and 7.9 over SFT on reasoning tasks.
-
Reinforcement Learning from Rich Feedback with Distributional DAgger
DistIL applies distributional DAgger with forward cross-entropy to achieve monotonic policy improvement and better Pass@N from rich feedback in RL for reasoning tasks.
-
Learning from Language Feedback via Variational Policy Distillation
VPD frames language feedback learning as variational EM so the teacher policy refines itself via trust-region updates on outcomes while the student learns dense token distributions on its own rollouts, outperforming fixed-teacher baselines on reasoning and code tasks.
-
Learn where to Click from Yourself: On-Policy Self-Distillation for GUI Grounding
GUI-SD introduces on-policy self-distillation with visually enriched privileged context and entropy-guided weighting, outperforming GRPO and naive OPSD on six GUI grounding benchmarks while improving training efficiency.
-
Don't Let Gains FADE: Breaking Down Policy Gradient Weights in RL
FADE is a self-adapting advantage for policy-gradient RL that reads training dynamics to balance positive/negative gradient mass and difficulty focus, yielding faster peak performance and better accuracy-diversity trade-offs than static baselines on LLM reasoning benchmarks.
-
REVES: REvision and VErification--Augmented Training for Test-Time Scaling
REVES augments LLM post-training by decoupling revision and verification signals from successful multi-step trajectories, reporting +6.5 point gains on LiveCodeBench over RL baselines.
-
Agent Explorative Policy Optimization for Multimodal Agentic Reasoning
AXPO addresses the Thinking-Acting Gap in agentic RL training by targeted resampling of tool calls in all-wrong subgroups, delivering +1.8pp gains over GRPO on nine multimodal benchmarks with an 8B model beating a 32B baseline on Pass@4.
-
Nudging Beyond the Comfort Zone: Efficient Strategy-Guided Exploration for RLVR
NudgeRL conditions RLVR rollouts on strategy-level contexts to drive diverse trajectories and applies an inter/intra-context reward decomposition plus distillation objective, outperforming GRPO and oracle baselines on math benchmarks.
-
Rollout Pass-Rate Control: Steering Binary-Reward RL Toward Its Most Informative Regime
Prefix Sampling replays self-generated trajectory prefixes to control rollout pass rates near 50% in binary-reward RL, delivering wall-clock speedups and modest performance gains on SWE-bench Verified and AIME tasks.
-
Uni-OPD: Unifying On-Policy Distillation with a Dual-Perspective Recipe
Uni-OPD unifies on-policy distillation across LLMs and MLLMs with dual-perspective strategies that promote student exploration and enforce order-consistent teacher supervision based on outcome rewards.
-
Scaling Self-Play with Self-Guidance
SGS adds self-guidance to LLM self-play for Lean4 theorem proving, surpassing RL baselines and enabling a 7B model to outperform a 671B model after 200 rounds.
-
AutoOR: Scalably Post-training LLMs to Autoformalize Operations Research Problems
AutoOR uses synthetic data generation and RL post-training with solver feedback to enable 8B LLMs to autoformalize linear, mixed-integer, and non-linear OR problems, matching larger models on benchmarks.
-
PolicyLong: Towards On-Policy Context Extension
PolicyLong shifts long-context data synthesis to an on-policy loop that re-screens contexts using the evolving model's entropy landscape, producing a self-curriculum that outperforms static offline baselines with larger gains at longer lengths.
-
LLMs learn scientific taste from institutional traces across the social sciences
Fine-tuned LLMs trained on social science publication records outperform experts and frontier models at judging which research pitches deserve attention.
-
Hide to Guide: Learning via Semantic Masking
SMEPO applies fine-grained semantic masking to expert guidance in RLVR, turning hard problems into fill-in-the-blank tasks while preserving structure, yielding up to 3.2 point accuracy gains and 4.2x faster training.
-
DenoiseRL: Bootstrapping Reasoning Models to Recover from Noisy Prefixes
DenoiseRL optimizes recovery from noisy prefixes in weak-model reasoning failures to improve performance and self-correction on math and general reasoning benchmarks without external supervision.