Search papers, labs, and topics across Lattice.
4
0
7
3
StepGuard slashes attack success rates by over 77% while maintaining nearly all utility, setting a new standard for safety in LLM-based agents.
Self-evolving agents can now learn more efficiently in resource-constrained environments by explicitly structuring experience into a tool graph memory that facilitates planning and tool reuse.
Reasoning SFT doesn't just memorize, it generalizes鈥攂ut only if you train it long enough, feed it good data, and use a capable model, and even then, reasoning gains come at the cost of safety.
Current LLM safety evaluations miss the mark: ATBench reveals how risks in realistic, multi-step agent interactions emerge over time, challenging even the strongest models.
