Search papers, labs, and topics across Lattice.
4
0
7
LLMs still struggle to reason in context when cultural and linguistic nuances are involved, achieving only 44% accuracy on a new grounded benchmark spanning 14 languages.
Forget about re-balancing losses – gradient geometry is the key to unlearning in LLMs without sacrificing retention.
Current memory systems, despite their complexity, are surprisingly worse than naive RAG when applied to continuous lifelogging scenarios, revealing a critical need for better context preservation.
PPO can be made sample-efficient and stable for long-horizon reasoning in LLMs by treating the problem as a sequence-level contextual bandit, sidestepping the need for computationally expensive multi-sampling.
