All Articles
16196 articles total
At Waymo, an AI project isn't ready until its evals are — not when the model performs well
Waymo outlines its 'eval-centric development' approach, emphasizing continuous, rigorous evaluation of AI models as a prerequisite for production deployment.
Data Quality Profiling at Scale with Progressive Sampling: A Benchmark for Data-Centric AI Pipelines
A benchmark study on data quality profiling showing that simple random uniform sampling often outperforms complex proxy-guided sampling at scale.
Raven: High-Recall Sequence Modeling with Sparse Memory Routing
Introduction of Raven, a linear-time sequence model using sparse memory routing to improve long-context recall compared to SSMs and sliding-window attention.
Rethinking Likelihood distributions: Student's t Likelihood Boosts Bayesian Neural Network Performance
Research demonstrating that using Student's t-likelihood distributions instead of Gaussian distributions improves performance in Bayesian Neural Networks.
MARS: Multi-Agent Re-ranking for Repeat-Order Food Delivery Recommendation
Presentation of MARS, a multi-agent re-ranking framework for food delivery recommendations combining LLMs with collaborative retrieval.
From Dyad to Triad: Eliciting XAI Requirements in Stroke Rehabilitation
A study on a video-based scaffolding protocol to elicit explainable AI (XAI) requirements from stroke survivors with communication disorders.
Emergent Latent-State Computation under Stochastic Volatility
Research using stochastic volatility as a benchmark for mechanistic interpretability, revealing how Transformers represent latent states under noisy observations.
The coolest use for the Vision Pro
Discussion regarding the most creative or useful applications of the Apple Vision Pro headset.
SalesPatriot (YC W25) Is Hiring FDEs
SalesPatriot, a Y Combinator W25 company, is currently hiring Forward Deployed Engineers.
Theo Conjecture solves 35-year-old math problem, finds a term no one predicted
The Theo Conjecture has been solved after 35 years, revealing an unexpected mathematical term.
A field day for Gleam's language server – Gleam v1.18.0 release
Gleam releases version 1.18.0, featuring significant updates to its language server.
Thinking Machines co-founder Lilian Weng left the company citing health reasons, then joined OpenAI
Lilian Weng has joined OpenAI after leaving Thinking Machines due to health reasons.
Winamp aims for a comeback with a new music player powered by Deezer
Winamp is attempting a brand comeback with a new music player powered by Deezer.
xAI’s last-minute scramble to stop Minnesota’s anti-nudification app law
xAI is suing Minnesota over a law targeting 'nudification' apps, claiming it violates the First Amendment.
Cyberpunk 2077 packs a lot of fun into its discounted $20 price
Cyberpunk 2077 is currently available at a significant discount across multiple gaming platforms.
Who wins and who loses after US bans foreign robots?
Analysis of the impact of US bans on foreign-made robots on the domestic robotics industry.
Comcast store punished low sales by smashing pies in workers' faces, lawsuit claims
Comcast faces a lawsuit alleging that store employees were punished with pies smashed in their faces.
Aurora DSQL: Scalable, Multi-Region OLTP
Aurora DSQL is presented as a scalable, multi-region Online Transactional Processing (OLTP) database system.
A Trampoline
A discussion or article titled 'A Trampoline' is mentioned, likely within a technical context but lacking detailed content.
Xbox revenue drops 10 percent as Microsoft’s cloud and AI business surges
Microsoft's Xbox revenue and hardware sales have declined, while its cloud and AI business segments have seen significant growth.
