8 papers
Rubric Dropout: A Simple Way to Mitigate Reward Hacking in Rubric-as-Reward RL
Minglai Yang, Xinyu Guo, Utkarsh Tyagi +6
Reinforcement learning against rubrics, lists of criteria graded by an LLM judge, has become a standard way to post-train language models on tasks with no deterministic answer. The…
Model or Harness? An Interaction-Centric Taxonomy for Localizing Agent Failures
Harsh Raj, Vipul Gupta, Anas Mahmoud +4
Existing evaluations often reduce agent failures to system-level outcomes, obscuring where the fault originated and which intervention would improve the agent system. This creates…
CRAFT: Clustering Rubrics to Diagnose Weak LLM Capabilities and Generate Targeted Fine-Tuning Data
Vipul Gupta, Zihao Wang, Razvan-Gabriel Dumitru +3
Evaluations should do more than measure a models current performance. They should tell us what to fix for the next model iteration and provide a way to generate targeted post train…
Rubric-Guided Self-Distillation: Post-Training Without Rubric Verifiers
MohammadHossein Rezaei, Anas Mahmoud, Zihao Wang +6
Rubrics have emerged as an alternative to RLVR in open-ended domains where a single ground-truth final answer is not available. Existing rubric-based training methods rely on an LL…
CopySpec: Accelerating LLMs with Speculative Copy-and-Paste Without Compromising Quality
Razvan-Gabriel Dumitru, Minglai Yang, Vikas Yadav +1
We introduce CopySpec, a simple yet effective technique to tackle the inefficiencies LLMs face when generating responses that closely resemble previous outputs or responses that ca…
ConciseRL: Conciseness-Guided Reinforcement Learning for Efficient Reasoning Models
Razvan-Gabriel Dumitru, Darius Peteleaza, Vikas Yadav +1
Large language models excel at complex tasks by breaking down problems into structured reasoning steps. However, reasoning traces often extend beyond reaching a correct answer, cau…