adaptive importance sampling fixes quantized rl training
a new method corrects policy gradient bias from low-precision rollouts in llm reinforcement learning, preventing training collapse.
aitag machine learning
summaries filed under machine learning
granite multilingual r2: compact 97m model tops sub-100m retrieval
ibm releases two apache 2.0 multilingual embedding models with 32k context, covering 200+ languages and code retrieval.
aiai confidence alignment speeds human decision learning
study shows that matching ai confidence to human self-confidence helps people learn faster when making decisions with ai assistance.
aiasynchronous batching speeds up llm inference
separating cpu and gpu work with cuda streams and events reduces idle time, boosting throughput by up to 24%.
ai5 small language models for agentic tool calling
five compact open-weight language models that support structured tool calling for agentic ai workflows.
aiocean concept bottleneck model explains heat predictions
a new model uses physics-based concepts to make ocean heat forecasts interpretable, revealing the drivers behind predictions.
ailearning when to act for safe, communication-efficient rl
a new method learns both control and when to communicate, using a safety shield to reduce sampling while maintaining stability.
aiquide: a single score for quantized model efficiency
a new metric called quide collapses compression, accuracy, and latency into one score to find optimal quantization levels for neural networks.
ai10 github repos to learn self-hosting
a curated list of github repositories that teach self-hosting skills from discovery to deployment, monitoring, and secure access.
aipytorch 2.12: faster linalg, unified graphs, and mx export
pytorch 2.12 brings up to 100x faster batched eigendecomposition, a device-agnostic graph api, and support for exporting microscaling quantized models.
aiadaptive steering for discrete diffusion language models
uniform control schedules hurt text quality in discrete diffusion models, but a new adaptive scheduler improves multi-attribute steering by aligning interventions with each attribute's unique denoising timeline.
aivertex-softmax tightens transformer verification
new method finds exact softmax bounds from score intervals, reducing slack in transformer robustness checks.
aibalora adds bayesian uncertainty to lora fine-tuning
balora extends lora with bayesian adaptation, improving accuracy and providing uncertainty estimates for large model fine-tuning.
aipython scripts for time series analysis
five reusable python scripts handle common time series tasks like resampling, anomaly detection, decomposition, forecasting, and multi-series comparison.
aiexecutorch labs for arm edge ai
arm releases hands-on jupyter labs to deploy pytorch models on edge cpus and npus using executorch.
aipolars vs pandas: real-world performance deep dive
a comparison of polars and pandas on three data tasks shows polars is faster and uses less memory due to lazy evaluation and parallelism.
ai10 github repos to learn fastapi
a curated list of ten github repositories that help developers learn fastapi through templates, examples, auth tools, microservices, and ml projects.
airl-kirigami speeds inverse design of shape-shifting structures
a new framework combines flow matching and reinforcement learning to quickly generate valid kirigami cut patterns for target shapes.
aipathboost: gradient boosting learns graph features from paths
pathboost is a gradient tree boosting method for graph-level prediction that learns path-based features directly from graph structure, outperforming graph neural networks on half of benchmark datasets.
aibuild an ai lms with ollama, fastapi, and react
a step-by-step guide to creating a learning management system that adapts to each learner using local ai models.
ai