Search papers, labs, and topics across Lattice.
Renmin University of China
7
0
9
3
AREX outperforms traditional models by recursively refining answers through a novel self-improvement mechanism that leverages intermediate verification.
Explicit evidence graphs in VeriGraph enable LLMs to achieve 87.61% claim grounding, transforming how we verify AI-generated conclusions.
Arbor's innovative approach to autonomous research enables a cumulative learning process that outperforms existing models by over 2.5 times in real-world tasks.
LLM agents are shockingly vulnerable to multi-stage "trojan" attacks that inject malicious instructions into their workspace, achieving near-perfect success rates where standard prompt injection defenses fail.
Scaling out peer agents with a shared reasoning hub, AgentFugue, unlocks a new dimension of capability gains in long-horizon tasks, proving that collective reasoning is more than just parallel compute.
LLMs can now recall relevant information from long interaction histories without retraining, thanks to a state-adaptive memory framework that uses lightweight cues to reconstruct distant information on demand.
A small, outcome-driven proxy model can outperform complex indexing methods for LLM memory retrieval, offering a more efficient and scalable solution for long-term tasks.
