EvoRubrics introduces adversarial co-evolution between policy LLM and rubric generator to produce dynamic, adaptive rewards for reinforcement learning on open-ended tasks.
hub Mixed citations
Breaking the exploration bottleneck: Rubric-scaffolded reinforcement learning for general llm reasoning.arXiv preprint arXiv:2508.16949
Mixed citation behavior. Most common role is background (60%).
17
Pith papers citing it
Background
60% of classified citations
hub tools
citation-role summary
background 5
citation-polarity summary
years
2026 17roles
background 5representative citing papers
rDPO uses offline-built rubrics to generate on-policy preference data for DPO, raising benchmark scores in visual tasks over outcome-based filtering and style baselines.
ARL-RR alternates optimization over rubric meta-classes with dynamic selection to avoid fixed scalarization, outperforming baselines on HealthBench.
T2R proposes a self-improving Teacher and competition-guided Reasoner with case-wise rewards to provide more effective supervision for CoT optimization in CXR VQA.
DR-rubric is a two-stage framework using iterative agentic search to generate atomic verifiable constraints for GRPO-based RL, achieving competitive performance on 6 benchmarks with 1K-3K examples via bootstrap or frontier-model rubrics.
RUBRIC-ARROW is an alternating rubric generator and judge framework that uses probability-based scoring and pairwise preferences to improve pointwise reward modeling accuracy for LLM post-training in non-verifiable domains.
Rubric-based RL verifiers can be gamed via partial criterion satisfaction and implicit-to-explicit tricks, yielding proxy gains that do not improve quality under rubric-free judges; stronger verifiers reduce but do not eliminate the mismatch.
Rubric-grounded RL with LLM judges on document-derived criteria raises Llama-3.1-8B normalized reward to 71.7% on held-out rubrics and improves performance on GSM8K, MATH, and GPQA benchmarks.
POP bootstraps post-training signals for open-ended LLM tasks by synthesizing rubrics during self-play on pretraining corpus, yielding performance gains on Qwen-2.5-7B across healthcare QA, creative writing, and instruction following.
RTT bridges response-level rubrics to token-level rewards via a relevance discriminator and intra-sample group normalization, yielding higher instruction and rubric accuracy than baselines.
QUBRIC co-designs queries and rubrics via teacher key points, contrastive generation, and learnability filtering to support GRPO training, yielding +5.5 on ArenaHard and +6.3 average transfer to legal/moral/narrative benchmarks.
TrOPD stabilizes on-policy distillation for LLMs with trust-region learning, outlier estimation, and off-policy guidance, outperforming prior OPD methods on reasoning and code benchmarks.
RLR³ extends RLVR to criterion-level rubric verification via dual execution paths, minimal exposure masking, hierarchical aggregation, and saturation mitigation, delivering 4.7-point gains over base on 15 benchmarks with Qwen3-VL-30B-A3B.
ARES generates 100K rubric-annotated QA instances from raw documents and demonstrates superior rubric-based RL performance over baselines on open-ended benchmarks.
Mid-training LLMs on self-generated diverse reasoning paths improves subsequent RL performance on mathematical benchmarks and OOD tasks.
LLM post-training is unified as off-policy or on-policy interventions that expand support for useful behaviors, reshape policies within reachable states, or consolidate behavior across training stages.
Randomly masking a proposer's output vocabulary during training and generation sustains curriculum diversity and improves solver accuracy by +4.4 points at 8B in LLM co-evolution.
citing papers explorer
-
EvoRubrics: Dynamic Rubrics as Rewards via Adversarial Co-Evolution for LLM Reinforcement Learning
EvoRubrics introduces adversarial co-evolution between policy LLM and rubric generator to produce dynamic, adaptive rewards for reinforcement learning on open-ended tasks.
-
Visual Preference Optimization with Rubric Rewards
rDPO uses offline-built rubrics to generate on-policy preference data for DPO, raising benchmark scores in visual tasks over outcome-based filtering and style baselines.
-
Alternating Reinforcement Learning with Contextual Rubric Rewards: Beyond the Scalarization Strategy
ARL-RR alternates optimization over rubric meta-classes with dynamic selection to avoid fixed scalarization, outperforming baselines on HealthBench.
-
Teach-to-Reason: Competition-Guided Reasoning with a Self-Improving Teacher
T2R proposes a self-improving Teacher and competition-guided Reasoner with case-wise rewards to provide more effective supervision for CoT optimization in CXR VQA.
-
Deep Research as Rubric for Reinforcement Learning
DR-rubric is a two-stage framework using iterative agentic search to generate atomic verifiable constraints for GRPO-based RL, achieving competitive performance on 6 benchmarks with 1K-3K examples via bootstrap or frontier-model rubrics.
-
RUBRIC-ARROW: Alternating Pointwise Rubric Reward Modeling for LLM Post-training in Non-verifiable Domains
RUBRIC-ARROW is an alternating rubric generator and judge framework that uses probability-based scoring and pairwise preferences to improve pointwise reward modeling accuracy for LLM post-training in non-verifiable domains.
-
Reward Hacking in Rubric-Based Reinforcement Learning
Rubric-based RL verifiers can be gamed via partial criterion satisfaction and implicit-to-explicit tricks, yielding proxy gains that do not improve quality under rubric-free judges; stronger verifiers reduce but do not eliminate the mismatch.
-
Rubric-Grounded RL: Structured Judge Rewards for Generalizable Reasoning
Rubric-grounded RL with LLM judges on document-derived criteria raises Llama-3.1-8B normalized reward to 71.7% on held-out rubrics and improves performance on GSM8K, MATH, and GPQA benchmarks.
-
Bootstrapping Post-training Signals for Open-ended Tasks via Rubric-based Self-play on Pre-training Text
POP bootstraps post-training signals for open-ended LLM tasks by synthesizing rubrics during self-play on pretraining corpus, yielding performance gains on Qwen-2.5-7B across healthcare QA, creative writing, and instruction following.
-
Rubrics to Tokens: Bridging Response-level Rubrics and Token-level Rewards in Instruction Following Tasks
RTT bridges response-level rubrics to token-level rewards via a relevance discriminator and intra-sample group normalization, yielding higher instruction and rubric accuracy than baselines.
-
QUBRIC: Co-Designing Queries and Rubrics for RL Beyond Verifiable Rewards
QUBRIC co-designs queries and rubrics via teacher key points, contrastive generation, and learnability filtering to support GRPO training, yielding +5.5 on ArenaHard and +6.3 average transfer to legal/moral/narrative benchmarks.
-
Trust Region On-Policy Distillation
TrOPD stabilizes on-policy distillation for LLMs with trust-region learning, outlier estimation, and off-policy guidance, outperforming prior OPD methods on reasoning and code benchmarks.
-
Reinforcement Learning with Robust Rubric Rewards
RLR³ extends RLVR to criterion-level rubric verification via dual execution paths, minimal exposure masking, hierarchical aggregation, and saturation mitigation, delivering 4.7-point gains over base on 15 benchmarks with Qwen3-VL-30B-A3B.
-
ARES: Automated Rubric Synthesis for Scalable LLM Reinforcement Learning
ARES generates 100K rubric-annotated QA instances from raw documents and demonstrates superior rubric-based RL performance over baselines on open-ended benchmarks.
-
Mid-Training with Self-Generated Data Improves Reinforcement Learning in Language Models
Mid-training LLMs on self-generated diverse reasoning paths improves subsequent RL performance on mathematical benchmarks and OOD tasks.
-
Large Language Model Post-Training: A Unified View of Off-Policy and On-Policy Learning
LLM post-training is unified as off-policy or on-policy interventions that expand support for useful behaviors, reshape policies within reachable states, or consolidate behavior across training stages.
-
Vocabulary Dropout for Curriculum Diversity in LLM Co-Evolution
Randomly masking a proposer's output vocabulary during training and generation sustains curriculum diversity and improves solver accuracy by +4.4 points at 8B in LLM co-evolution.