Ye Yu

LLM agents can now build lasting, reusable "competence" for specific knowledge sources, moving beyond simple retrieval to deep, adaptive understanding.

OptiPrime is introduced, a protocol-hardware co-optimization framework for efficient private DNN inference that features a novel HE protocol for convolutions that substantially reduces the number of transmitted output ciphertexts and mitigates the network communication bottleneck.

ROSETTA is proposed, a hybrid CKKS/TFHE framework that overcomes inefficiency in evaluating nonlinear operations, which incur substantial overhead and dominate the decode stage and achieves up to $4.8\times$ Softmax speedup and $1.5$--$2.1\times$ end-to-end speedup over the SOTA framework CacheMir.

CARD achieves a level of realism in credit card discussions that outstrips traditional simulation methods, revealing the power of structured conversational planning.

DDR-Net outperforms existing dehazing methods by effectively refining features and enhancing structural representation, setting a new standard for single-image dehazing.

Switching between autoregressive and diffusion modes allows Nemotron-Labs-Diffusion to achieve unprecedented throughput and efficiency in language modeling.

Latent reasoning can now be validated in real-time, with a novel method that boosts accuracy by over 16 points on critical benchmarks.

Ditch the fixed interface: DiffMAS unlocks surprisingly large gains in multi-agent reasoning by jointly optimizing latent communication, outperforming text-based and prior latent methods by a wide margin.