11 papers · 1 filter
One Adapter Pair per Model: A Universal Activation Interface for Language Models
Su-Hyeon Kim, Jiwan Mun, Yo-Sub Han
Activation-based tools are usually tied to one model's native hidden space, requiring probes, sparse autoencoders, and natural-language interpreters to be rebuilt or rediscovered f…
STAB: Specification-driven Testing for Algorithmic Bottlenecks
Soohan Lim, Joonghyuk Hahn, Hyundong Jin +1
Evaluating the efficiency of algorithmic code requires test cases that expose runtime bottlenecks. Previous methods generate efficiency test cases either by increasing input size o…
Cross-Family Universality of Behavioral Axes via Anchor-Projected Representations
Su-Hyeon Kim, Yo-Sub Han
Large language models from different families use different hidden dimensions, tokenizers, and training procedures, making behavioral directions difficult to compare or transfer ac…
ContractEval: A Benchmark for Evaluating Contract-Satisfying Assertions in Code Generation
Soohan Lim, Joonghyuk Hahn, Hyunwoo Park +2
Current code generation evaluation measures functional correctness on well-formed inputs that satisfy all input preconditions. This paradigm has a critical limitation: task descrip…
Detection of LLM-Paraphrased Code and Identification of the Responsible LLM Using Coding Style Features
Shinwoo Park, Hyundong Jin, Jeong-won Cha +1
Recent progress in large language models (LLMs) for code generation has raised serious concerns about intellectual property protection. Malicious users can exploit LLMs to produce…
How Does the Thinking Step Influence Model Safety? An Entropy-based Safety Reminder for LRMs
Su-Hyeon Kim, Hyundong Jin, Yejin Lee +1
Large Reasoning Models (LRMs) achieve remarkable success through explicit thinking steps, yet the thinking steps introduce a novel risk by potentially amplifying unsafe behaviors.…