activity
20172026
most citedNeural Text Generation with Unlikelihood Training

241 citations · 538 across the 42 of their papers we have counts for

collaborators
Showing cs.CLShow all

24 papers · 1 filter

cs.CL2026

On the limits and opportunities of AI reviewers: Reviewing the reviews of Nature-family papers with 45 expert scientists

Seungone Kim, Dongkeun Yoon, Kiril Gashteovski +55

With the advancement of AI capabilities, AI reviewers are beginning to be deployed in scientific peer review, yet their capability and credibility remain in question: many scientis…

cs.CL2026

Soohak: A Mathematician-Curated Benchmark for Evaluating Research-level Math Capabilities of LLMs

Guijin Son, Seungone Kim, Catherine Arnett +73

Following the recent achievement of gold-medal performance on the IMO by frontier LLMs, the community is searching for the next meaningful and challenging target for measuring LLM…

cs.CL2026

Argument Reconstruction as Supervision for Critical Thinking in LLMs

Hyun Ryu, Gyouk Chu, Gregor Betz +3

To think critically about arguments, human learners are trained to identify, reconstruct, and evaluate arguments. Argument reconstruction is especially important because it makes a…

cs.CL20251 cited

RefineBench: Evaluating Refinement Capability of Language Models via Checklists

Young-Jun Lee, Seungone Kim, Byung-Kwan Lee +6

Can language models (LMs) self-refine their own responses? This question is increasingly relevant as a wide range of real-world user interactions involve refinement requests. Howev…

cs.CL2025

Agentic-R1: Distilled Dual-Strategy Reasoning

Weihua Du, Pranjal Aggarwal, Sean Welleck +1

Current long chain-of-thought (long-CoT) models excel at mathematical reasoning but rely on slow and error-prone natural language traces. Tool-augmented agents address arithmetic v…

cs.CL20251 cited

OptimalThinkingBench: Evaluating Over and Underthinking in LLMs

Pranjal Aggarwal, Seungone Kim, Jack Lanchantin +4

Thinking LLMs solve complex tasks at the expense of increased compute and overthinking on simpler problems, while non-thinking LLMs are faster and cheaper but underthink on harder…