Leap
PulseAugur coverage of Leap — every cluster mentioning Leap across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
New RL framework LEAP optimizes GPU kernel generation
Researchers have developed LEAP, a new reinforcement learning framework designed for generating GPU kernels. This framework addresses challenges like sparse rewards and long compilation times by using a Difficulty-Condi…
-
openSUSE Planet: AI assistant, Cyber Resilience Act, and ODF mandate covered
The openSUSE Planet roundup for the week highlights several key developments in the open-source and technology policy landscape. A keynote at oSC26 focused on sovereign open-source assurance and the implications of the …
-
New LEAP curriculum boosts Vision Transformer distillation efficiency
Researchers from the University of Oxford have introduced LEAP, a novel training curriculum designed to improve the efficiency of knowledge distillation for Vision Transformers (ViTs). LEAP utilizes a progressive approa…
-
New LEAP Method Enables Soft Robots to Adapt to Damage in Seconds
Researchers have developed a method called LEAP (Learned Ensemble Adaptation Proprioception) that enables soft robots to adapt to catastrophic damage in under a minute. This technique leverages architected materials, wh…
-
New robot policy models enhance action generation and efficiency
Researchers have developed new methods for robot policy learning that improve efficiency and accuracy in action generation. LeaP, a learnable source prior, optimizes the starting point for action generation by condition…
-
Google's LEAP framework grounds LLMs in Lean compiler
Google has developed a new research framework called LEAP that utilizes a general-purpose LLM within an agentic scaffold. This system grounds each step in the Lean compiler and refines its process through feedback from …
-
New LEAP Protocol Prevents Data Leakage in Early Warning Models
Researchers have developed a new protocol called LEAP (Leakage-Excluded Early-Availability Protocol) to address temporal leakage in early-warning models for Learning Management Systems (LMS). This protocol ensures that …
-
LLM-driven framework accelerates perovskite additive discovery
Researchers have developed LEAP, a closed-loop framework that uses a domain-specific large language model combined with active learning to discover additives for perovskite solar cells. This LLM is trained to extract kn…
-
Researchers explore weight decay, in-context learning, and acceleration for Transformer models
Researchers have developed several new methods to improve the efficiency and theoretical understanding of Transformer models. One paper provides a functional-analytic characterization of weight decay, demonstrating its …