Search papers, labs, and topics across Lattice.
AI Innovation, DeCoDE Lab, Red Hat
6
0
11
Length-adjusted tail entropy can serve as a powerful signal for solution quality, enabling significant performance boosts in inference-time scaling across complex domains.
Trading a fraction of inference compute for a threefold reduction in training costs, sGPO redefines efficiency in RLVR training.
Forward-Forward learning can finally compete with backpropagation on complex image tasks, thanks to a novel covariance-aware goodness function that captures crucial second-order feature dependencies.
You can detect prompt injection attacks in screenshot-based web agents with 8x speedup and no extra memory by looking for telltale visual "smoothness" and reversed text polarity.
Contact-aware reconstruction transforms how we achieve realistic human-scene interactions in 3D environments, correcting artifacts that have plagued previous methods.
AI can now provide real-time feedback on coding consistency in qualitative research, previously a manual and drift-prone process.
