Search papers, labs, and topics across Lattice.
7
0
10
Dynamo achieves a remarkable 5.6% average accuracy boost in visual reasoning tasks without retraining, revolutionizing how VLMs adapt in real-time.
PolicyAlign enables LLMs to adapt to rapidly changing safety policies without relying on expensive supervision data, achieving significant safety improvements across diverse applications.
Self-distillation can be transformed from mere imitation of a privileged distribution to a powerful tool for diagnosing and correcting specific reasoning failures in large language models.
Full-duplex dialogue systems are often mischaracterized, with many claiming capabilities they cannot deliver due to training limitations.
Conflict-aware filtering in GD^2PO boosts reinforcement learning efficiency by preventing negative signal cancellation from competing rewards.
By explicitly conditioning on the query, QGS achieves a 0.62% CTR increase in a major commercial search engine, proving that generative models can beat traditional deep learning baselines in search ranking when query context is properly handled.
Knowledge injection, reasoning supervision, and preference optimization can be combined to substantially improve semantic relevance judgment, outperforming even strong LLM baselines.
