kv cache as editable, composable notes
prefix caching fails when one field changes, but treating the kv cache as a notebook of memoized conclusions enables editing and composing precomputed skills.
aisummaries filed under research
prefix caching fails when one field changes, but treating the kv cache as a notebook of memoized conclusions enables editing and composing precomputed skills.
aigoogle deepmind and uk government prototype uses gemini to speed up householder planning decisions by 50%.
aigoogle research releases a vectorized dataset that turns high-resolution satellite maps into an actionable inventory of fine-scale woody features like hedgerows and copses across the uk.
aia new variational inference method uses splines to capture complex latent dependencies, moving beyond mean-field assumptions.
aiexport restrictions on claude fable 5 stem from a misinterpreted jailbreak that was actually a defensive code review request.
aia new algorithm tackles dynamic routing of queries to multiple embedding models under adversarial conditions and bandit feedback.
aia new method uses a small target dataset to train an audit model that improves prediction set coverage for pretrained classifiers facing unknown distribution shifts.
aia new definition of good explanations considers the listener's prior beliefs, highlighting why explaining large language model outputs is especially hard.
aia new framework combines retrieval with direct corpus interaction, letting agents pull relevant documents into a local workspace for flexible search and verification without full-corpus slowdowns.
aiaxios reports internal friction and export control issues at anthropic, with key staff meeting the commerce department amid model safety concerns.
airesearchers propose a behavioral measure of trust between ai agents using costly verification in a survival game, finding that larger models reduce verification with reliable teammates but differ in recovery after failures.
aia fully gpu-based method speeds up building neural emulators for hypersonic flows with physics-aware refinement and uncertainty quantification.
aieven with perfect learning, early stopping in diffusion models causes geometric drift toward a unique limiting distribution that acts as a low-pass filter on the data.
aia new method called lomc suppresses refusals in mixture-of-experts models using compact, support-gated edits that preserve general performance.
aia new framework uses user portraits and nested rollout to let large language models dynamically tailor dialogue policies without offline training.
aia study shows optimal transport with p-norm cost can recover rotation, translation, or scaling between source and target domains in 2d, enabling linear regression adaptation with scarce target data.
aia new framework extends decision trees to any bregman divergence loss, generalizing classic cart and enabling flexible, interpretable models for diverse data types.
aia transformer-based scheduling policy trained with deep reinforcement learning solves open shop scheduling problems and scales to large instances without retraining.
aillada.cpp is the first npu-aware framework for accelerating diffusion large language models on smartphones, using multi-block speculative decoding and other techniques.
aisimon willison explores methods to map sql query result columns back to their source table and column in sqlite, enabling richer data display in datasette.
ai