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.
hub Canonical reference
Learning what reinforcement learning can't: Interleaved online fine-tuning for hardest questions
Canonical reference. 100% of citing Pith papers cite this work as background.
20
Pith papers citing it
Background
100% of classified citations
hub tools
citation-role summary
background 6
citation-polarity summary
roles
background 6polarities
background 6representative citing papers
NPO uses a policy's own near-future checkpoint as auxiliary trajectories to maximize effective learning signal S = Q/V, improving performance from 57.88 to 63.15 on Qwen3-VL-8B-Instruct with GRPO while accelerating convergence.
PaIR-Drive runs IL and RL in parallel branches with a tree-structured sampler to reach 91.2 PDMS and 87.9 EPDMS on NAVSIM benchmarks while outperforming sequential RL fine-tuning and correcting some human errors.
Prefix-RFT blends SFT and RFT via prefix sampling from demonstrations to outperform standalone SFT, RFT, and mixed-policy baselines on math reasoning problems.
AnE combines Truth Anchor Expansion and Scaffold-Stripping to deliver 10.3% gains on eight multimodal reasoning benchmarks for MLLMs.
SORT turns all-wrong prompts into selective learning signals by weighting tokens more predictable under plan guidance from reference solutions, improving over GRPO on reasoning benchmarks especially for weaker models.
AIPO adds active multi-agent consultation (Verify, Knowledge, Reasoning agents) plus custom importance sampling to RLVR training so LLMs expand their reasoning boundary and then operate without the agents.
DoTS decouples SFT and RLVR training then synthesizes their task vectors at inference time to match integrated training results at ~3% compute cost.
LLMs show strong spatial generalization to unseen maps in shortest-path tasks but fail length scaling due to recursive instability, with data coverage setting hard limits.
Lightning OPD is an offline on-policy distillation method that matches standard OPD performance at 4x efficiency by enforcing teacher consistency between SFT and distillation phases.
HAPO adds a hindsight-anchored SSI operator with Thompson gating to GRPO-style RLVR, achieving asymptotic consistency that recovers unbiased on-policy gradients as the policy improves.
SPHINX generates synthetic visual puzzles for benchmarking LVLMs, where GPT-5 scores 51.1% and RLVR training improves both in-domain and external visual reasoning performance.
ViSurf unifies SFT and RLVR for LVLMs in one training stage by injecting ground-truth labels into rollouts and applying novel reward controls, outperforming standalone and two-stage baselines on diverse benchmarks.
EvoCoT uses self-generated and verified CoT trajectories in a two-stage curriculum to let LLMs learn from initially unsolved hard problems in RLVR settings.
LoRR augments preference optimization methods like DPO with high-replay training, periodic resets to initial data/policy, and a hybrid objective to improve sample efficiency and reduce primacy bias on math and reasoning tasks.
BV-Blend blends prompt-local and semantic-cluster historical reward statistics via SEM-derived weights to stabilize critic-free RL advantage estimation.
Outcome-level RL with binary or composite rewards improves compositional generalization over supervised fine-tuning by avoiding overfitting to frequent training patterns.
EasyVideoR1 delivers an optimized RL pipeline for video understanding in large vision-language models, achieving 1.47x throughput gains and aligned results on 22 benchmarks.
citing papers explorer
-
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.
-
Near-Future Policy Optimization
NPO uses a policy's own near-future checkpoint as auxiliary trajectories to maximize effective learning signal S = Q/V, improving performance from 57.88 to 63.15 on Qwen3-VL-8B-Instruct with GRPO while accelerating convergence.
-
Fine-tuning is Not Enough: A Parallel Framework for Collaborative Imitation and Reinforcement Learning in End-to-end Autonomous Driving
PaIR-Drive runs IL and RL in parallel branches with a tree-structured sampler to reach 91.2 PDMS and 87.9 EPDMS on NAVSIM benchmarks while outperforming sequential RL fine-tuning and correcting some human errors.
-
Blending Supervised and Reinforcement Fine-Tuning with Prefix Sampling
Prefix-RFT blends SFT and RFT via prefix sampling from demonstrations to outperform standalone SFT, RFT, and mixed-policy baselines on math reasoning problems.
-
AnE: Pushing the Reasoning Frontier of Multimodal LLMs via Anchor Evolution
AnE combines Truth Anchor Expansion and Scaffold-Stripping to deliver 10.3% gains on eight multimodal reasoning benchmarks for MLLMs.
-
Selective Off-Policy Reference Tuning with Plan Guidance
SORT turns all-wrong prompts into selective learning signals by weighting tokens more predictable under plan guidance from reference solutions, improving over GRPO on reasoning benchmarks especially for weaker models.
-
AIPO: Learning to Reason from Active Interaction
AIPO adds active multi-agent consultation (Verify, Knowledge, Reasoning agents) plus custom importance sampling to RLVR training so LLMs expand their reasoning boundary and then operate without the agents.
-
Decouple before Integration: Test-time Synthesis of SFT and RLVR Task Vectors
DoTS decouples SFT and RLVR training then synthesizes their task vectors at inference time to match integrated training results at ~3% compute cost.
-
Generalization in LLM Problem Solving: The Case of the Shortest Path
LLMs show strong spatial generalization to unseen maps in shortest-path tasks but fail length scaling due to recursive instability, with data coverage setting hard limits.
-
Lightning OPD: Efficient Post-Training for Large Reasoning Models with Offline On-Policy Distillation
Lightning OPD is an offline on-policy distillation method that matches standard OPD performance at 4x efficiency by enforcing teacher consistency between SFT and distillation phases.
-
Hindsight-Anchored Policy Optimization: Turning Failure into Feedback in Sparse Reward Settings
HAPO adds a hindsight-anchored SSI operator with Thompson gating to GRPO-style RLVR, achieving asymptotic consistency that recovers unbiased on-policy gradients as the policy improves.
-
SPHINX: A Synthetic Environment for Visual Perception and Reasoning
SPHINX generates synthetic visual puzzles for benchmarking LVLMs, where GPT-5 scores 51.1% and RLVR training improves both in-domain and external visual reasoning performance.
-
ViSurf: Visual Supervised-and-Reinforcement Fine-Tuning for Large Vision-and-Language Models
ViSurf unifies SFT and RLVR for LVLMs in one training stage by injecting ground-truth labels into rollouts and applying novel reward controls, outperforming standalone and two-stage baselines on diverse benchmarks.
-
EvoCoT: Overcoming the Exploration Bottleneck in Reinforcement Learning
EvoCoT uses self-generated and verified CoT trajectories in a two-stage curriculum to let LLMs learn from initially unsolved hard problems in RLVR settings.
-
Sample-efficient LLM Optimization with Reset Replay
LoRR augments preference optimization methods like DPO with high-replay training, periodic resets to initial data/policy, and a hybrid objective to improve sample efficiency and reduce primacy bias on math and reasoning tasks.
-
BV-Blend: Uncertainty-Weighted Historical Baselines for Stable Critic-Free RL with Verifiable Rewards
BV-Blend blends prompt-local and semantic-cluster historical reward statistics via SEM-derived weights to stabilize critic-free RL advantage estimation.
-
Reinforcement Learning for Compositional Generalization with Outcome-Level Optimization
Outcome-level RL with binary or composite rewards improves compositional generalization over supervised fine-tuning by avoiding overfitting to frequent training patterns.
-
EasyVideoR1: Easier RL for Video Understanding
EasyVideoR1 delivers an optimized RL pipeline for video understanding in large vision-language models, achieving 1.47x throughput gains and aligned results on 22 benchmarks.
- Which Reasoning Trajectories Teach Students to Reason Better? A Simple Metric of Informative Alignment
- Position: The Hidden Costs and Measurement Gaps of Reinforcement Learning with Verifiable Rewards