7 papers
It Takes Two: Complementary Self-Distillation for Contextual Integrity in LLMs
Sangwoo Park, Woongyeong Yeo, Seanie Lee +6
Contextual Integrity (CI) defines privacy not merely as keeping information hidden, but as governing information flows according to the norms of a given context. As large language…
HINT-SD: Targeted Hindsight Self-Distillation for Long-Horizon Agents
Woongyeng Yeo, Yumin Choi, Taekyung Ki +1
Training long-horizon LLM agents with reinforcement learning is challenging because sparse outcome rewards reveal whether a task succeeds, but not which intermediate actions caused…
THINKSAFE: Self-Generated Safety Alignment for Reasoning Models
Seanie Lee, Sangwoo Park, Yumin Choi +6
Large reasoning models (LRMs) achieve remarkable performance by leveraging reinforcement learning (RL) on reasoning tasks to generate long chain-of-thought (CoT) reasoning. However…
PREPING: Building Agent Memory without Tasks
Yumin Choi, Sangwoo Park, Minki Kang +2
Agent memory is typically constructed either offline from curated demonstrations or online from post-deployment interactions. However, regardless of how it is built, an agent faces…
T-MAP: Red-Teaming LLM Agents with Trajectory-aware Evolutionary Search
Hyomin Lee, Sangwoo Park, Yumin Choi +3
While prior red-teaming efforts have focused on eliciting harmful text outputs from large language models (LLMs), such approaches fail to capture agent-specific vulnerabilities tha…
Multimodal Prompt Optimization: Why Not Leverage Multiple Modalities for MLLMs
Yumin Choi, Dongki Kim, Jinheon Baek +1
Large Language Models (LLMs) have shown remarkable success, and their multimodal expansions (MLLMs) further unlock capabilities spanning images, videos, and other modalities beyond…