recursive diffusion training converges to a smoothed limit
even 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.
topic
data science
localized multidirectional correction for refusal suppression in moe models
a new method called lomc suppresses refusals in mixture-of-experts models using compact, support-gated edits that preserve general performance.
optimal transport recovers geometric shifts for domain adaptation
a 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.
decision trees unified via bregman divergences
a new framework extends decision trees to any bregman divergence loss, generalizing classic cart and enabling flexible, interpretable models for diverse data types.
weekly brief: 2026-06-08-2026-06-14
spacex's record ipo, anthropic's model shutdown, and india's ai sovereignty debate lead a week of market shifts and regulatory moves.
daily brief: 2026-06-14
india debates ai self-reliance after anthropic blocks models, meta unwinds manus deal under pressure, and state attorneys general investigate openai.
daily brief: 2026-06-12
today's digest covers ai agent support, phishing lawsuits, evaluation tools, low-carbon computing, and more.
7 ways to fund your startup idea
seven funding options for startups, from bootstrapping to venture capital, with pros and cons for each.
run claude code with local models
how to connect claude code to ollama, lm studio, or llama.cpp for zero-cost, rate-limit-free coding sessions using local models.
3 numpy tricks for faster numerical code
learn vectorization, in-place operations, and memory views to speed up numpy code and reduce memory use.
epistemic uncertainty is not always reducible
new research shows that the standard definition of epistemic uncertainty as reducible by more data is inconsistent with its common mutual-information measure, and proposes a three-part taxonomy.
daily brief: 2026-06-11
ai agents struggle with scientific synthesis, deezer detects ai music, and doordash adds a chatbot for food orders.
build a minimal feature store from scratch
a hands-on guide to building a feature store with duckdb, parquet, redis, and fastapi, covering offline and online stores, materialization, and retrieval for both predictive ml and llm context.
conformal bayes under label shift
two approaches for conformal bayes under label shift: post-hoc calibration and in-training adaptation, both using importance weighting to restore target-domain coverage.
annealed entropic allocation for ranking and selection
a new annealed weighted soft-min framework improves sequential budget allocation in ranking and selection by smoothing the maximin objective and adding saddlepoint corrections.
topological features as survival data for testing and learning
a new method treats persistence diagrams as survival data, enabling hypothesis testing, effect sizes, and stable feature vectors for machine learning.
boltzmann margin gives knn near-exponential convergence
a new margin condition bridges the gap between polynomial and exponential rates for knn classifiers.
python scripts to automate pdf tasks
five python scripts that merge, split, extract, stamp, redact, and inventory pdfs from the command line.