Search papers, labs, and topics across Lattice.
6
0
7
13
Critic-R transforms agentic search by enabling retrieval models to learn from their own reasoning failures, significantly boosting answer accuracy without the need for gold-standard annotations.
Forget slow retrieval pipelines – GrepSeek trains agents to `grep` their way to answers directly from massive text corpora, achieving SOTA on open-domain QA.
Closing the gap between fixed retrieval systems and oracle performance in agentic search yields a +26.8% F1 improvement, demonstrating that jointly training the reasoning agent and retrieval system is a game-changer.
Stop evaluating AI systems in isolation: marketplace dynamics like user switching and early-adoption advantages critically shape real-world success.
On-device RAG gets a major efficiency boost: a unified model slashes context size by 90% without sacrificing accuracy.
Forget sparse rewards: SLATE uses LLMs to judge each reasoning step, slashing gradient variance and boosting performance on retrieval-augmented reasoning tasks, especially for smaller models.
