AI RESEARCH
Efficient Test-Time Finetuning of LLMs via Convex Reconstruction and Gradient Caching
arXiv CS.LG
•
ArXi:2605.30337v1 Announce Type: new Test-time finetuning (TTFT) is a rapidly evolving paradigm that adapts a language model to each prompt by retrieving related sequences, updating the model on them, and then evaluating the prompt. However, TTFT is only practical if it is fast: selection and finetuning both happen per query, making each a direct bottleneck. Existing methods trade speed for quality: fast retrieval is often redundant, while stronger diversity-aware selection adds prohibitive per-query cost. We