SafeMoE, a Mixture-of-Experts (MoE) framework that isolates unsafe knowledge into domain-specific Low-Rank Adapters (LoRA experts) trained exclusively on harmful corpora, is introduced, proposing a dialectical approach to alignment, positing that unsafe data encodes rich, domain specific knowledge critical for nuanced,...
M. Hashemzadeh, Jerry Huang, Minseon Kim et al.· arXiv.org· 0 citations
KV-streams is proposed, a plug-and-play strategy compatible with any compaction strategy that substantially increases throughput while showing no evidence of hindering performance, and is shown to be an efficient and lightweight plug-and-play addition to any post-training pipeline.
Emiliano Penaloza, Dane Malenfant, Dheeraj Vattikonda et al.· 0 citations
Retrospection-Only Fine-Tuning (ROFT), a minimal online procedure designed to isolate the effect of explanation-only training on subsequent behavior, is studied, showing that learning to explain can also improve learning to do and motivating further study of explanation-to-action transfer.
Jonathan Light, C. Cui, Jeonghye Kim et al.· 0 citations
ProgramDistill is introduced, a benchmark evaluating coding agents on features discovered through interaction with fully functional reference applications, and provides a scalable benchmark with controlled difficulty for evaluating and diagnosing coding agents, and a natural basis for future curriculum-based training.
Jeonghye Kim, Minseon Kim, Young Jin Kim et al.· 0 citations
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-trained exclusively via RL on around 1,500 SWE environments with synthetic tasks. A key ingredient for improving performance is an online task...
Minseon Kim, Zheng-Yan Shi, Emiliano Penaloza et al.· 2 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
