7 papers
KV-streams for Efficient Compaction in Agentic Reinforcement Learning
Emiliano Penaloza, Dane Malenfant, Dheeraj Vattikonda +15
Scaling the horizon of agentic LLMs is bottlenecked by the need to fit ever longer context traces in GPU memory. Context compaction has been the most popular mechanism to alleviate…
Shockingly Simple Self-retrospection Improves Agentic Models Without RL
Jonathan Light, Christopher Zhang Cui, Jeonghye Kim +7
People learn not only by repeating successful actions, but also by recounting and explaining their experiences, revising their understanding to guide future behavior. Can a languag…
FrogNano: Training a 4B Coding Agent via Online Task Synthesis
Minseon Kim, Zhengyan Shi, Emiliano Penaloza +14
We present FrogNano, a 4B coding agent designed to tackle software engineering (SWE) tasks efficiently and effectively, even under resource-constrained environments. It is post-tra…
Dialectics of Alignment: Harnessing Unsafe Knowledge for Dynamic Safety Routing
Maryam Hashemzadeh, Jerry Huang, Minseon Kim +2
The prevailing paradigm in large language model (LLM) alignment operates via erasure, filtering unsafe data or training models to strictly refuse harmful prompts. While effective a…
Learning to Extract Context for Context-Aware LLM Inference
Minseon Kim, Lucas Caccia, Zhengyan Shi +4
User prompts to large language models (LLMs) are often ambiguous or under-specified, and subtle contextual cues shaped by user intentions, prior knowledge, and risk factors strongl…
Gistify! Codebase-Level Understanding via Runtime Execution
Hyunji Lee, Minseon Kim, Chinmay Singh +10
As coding agents are increasingly deployed in large codebases, the need to automatically design challenging, codebase-level evaluation is central. We propose Gistify, a task where…
BugPilot: Complex Bug Generation for Efficient Learning of SWE Skills
Atharv Sonwane, Isadora White, Hyunji Lee +8
High quality bugs are key to training the next generation of language model based software engineering (SWE) agents. We introduce a novel method for synthetic generation of difficu…