RGSD distills rubric-conditioned teacher distributions into base policies token-by-token, matching GRPO rubric satisfaction on Qwen models with one rollout and zero verifier calls.
Primack, Summer Yue, and Chen Xing
10 Pith papers cite this work, alongside 1 external citations. Polarity classification is still indexing.
10
Pith papers citing it
1
external citations · OpenAlex
citation-role summary
dataset 1
citation-polarity summary
years
2026 10roles
dataset 1polarities
use dataset 1representative citing papers
A Gaussian information-gain metric in embedding space quantifies semantic progress in dialogues via uncertainty reduction and shows competitive agreement with human judgments on MT-Bench and UltraFeedback.
SCICONVBENCH is a new benchmark evaluating LLMs on multi-turn disambiguation and inconsistency resolution for task formulation in computational science, with frontier models reaching only 52.7% success on fluid mechanics disambiguation cases.
Prosa demonstrates that rubric-based binary scoring with multi-judge filtering yields full agreement on 16 LLM rankings across judges on Brazilian Portuguese chats, compared to only 7/16 under holistic scoring, while widening score gaps by 47%.
LMs systematically inflate expressed certainty during rewriting, affecting up to 75% of outputs with a 1.5-2x bias toward increasing rather than decreasing certainty, and the effect compounds over iterations.
Presents HarmAmp benchmark for multi-turn harm amplification in LLMs and TrajSafe proactive monitor that reduces harm while keeping low over-refusal and preserving capabilities.
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.
LoCar is a localization-aware evaluation framework for in-vehicle assistants that identifies unstable Korean honorific control and weaker performance on strategic metrics like clarification and proactivity in current LLMs.
Technical report announcing Ling-2.6 and Ring-2.6 models with hybrid linear attention, evolutionary CoT, and KPop RL for efficient agentic intelligence at scale.
Generative AI evaluation must shift from static benchmark scores to measuring sustained improvements in human capabilities within specific deployment contexts.
citing papers explorer
-
Rubric-Guided Self-Distillation: Post-Training Without Rubric Verifiers
RGSD distills rubric-conditioned teacher distributions into base policies token-by-token, matching GRPO rubric satisfaction on Qwen models with one rollout and zero verifier calls.
-
Measuring Semantic Progress in Multi-turn Dialogue via Information Gain
A Gaussian information-gain metric in embedding space quantifies semantic progress in dialogues via uncertainty reduction and shows competitive agreement with human judgments on MT-Bench and UltraFeedback.
-
SCICONVBENCH: Benchmarking LLMs on Multi-Turn Clarification for Task Formulation in Computational Science
SCICONVBENCH is a new benchmark evaluating LLMs on multi-turn disambiguation and inconsistency resolution for task formulation in computational science, with frontier models reaching only 52.7% success on fluid mechanics disambiguation cases.
-
Prosa: Rubric-Based Evaluation of LLMs on Real User Chats in Brazilian Portuguese
Prosa demonstrates that rubric-based binary scoring with multi-judge filtering yields full agreement on 16 LLM rankings across judges on Brazilian Portuguese chats, compared to only 7/16 under holistic scoring, while widening score gaps by 47%.
-
From `May' to `Is': Certainty Distortion in Language Model Rewriting
LMs systematically inflate expressed certainty during rewriting, affecting up to 75% of outputs with a 1.5-2x bias toward increasing rather than decreasing certainty, and the effect compounds over iterations.
-
Investigating and Alleviating Harm Amplification in LLM Interactions
Presents HarmAmp benchmark for multi-turn harm amplification in LLMs and TrajSafe proactive monitor that reduces harm while keeping low over-refusal and preserving capabilities.
-
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.
-
LoCar: Localization-Aware Evaluation of In-Vehicle Assistants through Fine-Grained Sociolinguistic Control
LoCar is a localization-aware evaluation framework for in-vehicle assistants that identifies unstable Korean honorific control and weaker performance on strategic metrics like clarification and proactivity in current LLMs.
-
Ling and Ring 2.6 Technical Report: Efficient and Instant Agentic Intelligence at Trillion-Parameter Scale
Technical report announcing Ling-2.6 and Ring-2.6 models with hybrid linear attention, evolutionary CoT, and KPop RL for efficient agentic intelligence at scale.
-
Benchmarked Yet Not Measured -- Generative AI Should be Evaluated Against Real-World Utility
Generative AI evaluation must shift from static benchmark scores to measuring sustained improvements in human capabilities within specific deployment contexts.