Proposes a method that optimizes policies using only verified negatives, even when positive verification is unavailable.
What is new.
What it means for your work.
Discover recent technology and evidence. Browse older material in All notes.
Changes to start with
Newly added notes across different industries
AgentRoboCritic · Action-Level Analysis for Evaluating Robot Control Capabilities
TideCache: Transparent Cache Management and Layer-wise Streaming for Hybrid LLM Serving
Virus-like particles enable efficient gene editing in pigs and poultry models
New technology notes
Originals from the past 14 days, ordered by when added to Blackdeck.
Multiple sub-tasks allow precise localization of the next sub-goal in sparse attention heads.
Uses explicit frame-to-frame correspondences to guide output generation and improve temporal consistency
Introduces functional transparency to represent model predictions as sums of univariate and bivariate functions.
Learned energy model CRAM solves reasoning tasks significantly faster than looping models
A sketching method reduces evaluation calls for context attribution by leveraging stability assumptions and backpropagation circuit differentiation
Future task success is evaluated by comparing runs with and without memory segments, enabling cost-aware memory selection.
AgentSSL proposes a method for MLE agents to combine and adapt established SSL techniques using unlabeled data.
Infineon successfully fabricated functional silicon carbide power devices using existing silicon manufacturing equipment on 300mm wafers.
Metal Sandbox enables isolated and scheduled command execution for GPU code generation agents on Apple Silicon
A framework is proposed that accounts for user adaptation in prompt strategies following LLM responses.
We propose a framework that estimates image creativity by predicting novelty, surprise, and arousal dimensions based on psychological principles.
TabPFN accurately predicts three key properties of energetic materials: impact sensitivity, decomposition temperature, and crystal density
Observes complementarity between cropping and compression, and proposes adaptive routing based on input image and question.
A specialized LLM-based system optimized for segment-level quality prediction and error-free segment detection in translation evaluation.