Search papers, labs, and topics across Lattice.
5
0
9
1
Adaptive multi-turn attacks can increase LLM defender vulnerability by over 14%, exposing critical weaknesses in existing safety benchmarks.
Manipulating 3D latent particles can generate novel scene configurations, revolutionizing how we approach object-centric representation in robotics.
A clever two-stage agent using smaller models can produce better, more substantive peer reviews than brute-force application of the largest LLMs.
Forget scraping the internet for physics Q&A: this work shows you can train LLMs to ace the International Physics Olympiad using only data generated from physics simulators.
Human-written solutions can actually *hurt* model performance on math problems, highlighting a critical gap between strategy usage and executability that Selective Strategy Retrieval (SSR) effectively bridges.
