> The principal concern of the farmer is risk. The crops cannot fail. If they… more
~research×
86 links
> Inspired by a behavior we’ve observed in real-world deployment, we evaluated… more
> New research from Visa Business and Economic Insights finds that $36 trillion… more
> I regret to inform you that a team of researchers, writing in a yet-to-be… more
> A meaningful text can be hidden inside another, completely different yet… more
> In response to struggling renters and frustrated first-time buyers,… more
> To check whether this behavior was more widespread, Ms. Van Belle and her… more
> The top 1% of firms spend $7.45K per employee per month. The top 10% spend… more
Firms that adopt AI grow headcount 10.2% over the two years following adoption,… more
> Abstract:Many everyday programming tasks resist clean rule-based… more
> This study analyzes the primary drivers of recent state-level trends in U.S.… more
> The quantitative aspirations of economists and financial analysts have for… more
> Laughter is an important, universal form of human non-linguistic vocal… more
> To understand why prompt injection happens, we need a way to measure what… more
> LLM-generated stories are a popular use case, but they show very low… more
> We use agentic misalignment as a case study to highlight some of the… more
> We introduce Natural Language Autoencoders (NLAs), an unsupervised method for… more
Abstract: more
> [...] We evaluate AgentFlow on TerminalBench-2 with Claude Opus 4.6 and on… more
> One way to tell GiveDirectly’s story is thus as a bellwether for… more
> We examine the economic impact of non-consumable visual cues through home… more
> Finnish nationwide cohort of all under-23-year-old gender-referred… more
> Abstract:Can a large language model (LLM) improve at code generation using… more
> [...] Frontier models readily generate detailed image descriptions and… more
> While Mixture-of-Experts (MoE) scales capacity via conditional computation,… more
> Project CETI (Cetacean Translation Initiative) has released two landmark… more
We’ve sent cash to nearly 1,500 pregnant women in rural Kenya to support safer… more
> "A generator is a parser of randomness." This perspective on generators for… more
> We've achieved 10x data efficiency with NanoGPT Slowrun within a few weeks.… more
> If you give a chimp a crystal, she might not give it back. more
> Prompt injection is a key problem in building reliable, long-running agents.… more
> In a two-year-long citywide experiment in Singapore, researchers divided… more
> We designed FDM-1, a foundation model for computer use. FDM-1 is trained on… more
> Following Uber-initiated fare increases, drivers make more money per trip… more
> We examine the extent to which security against a fully malicious server… more
> Using data from the nationally representative U.S. Health and Retirement… more
> We introduce Moshi, a speech-text foundation model and full-duplex spoken… more
> In our in-progress research, we discovered that AI tools didn’t reduce work,… more
> "Our radiation-hydrodynamic simulations track both how gas falls onto the… more
> A short introduction to RLHF and post-training focused on language models. more
> Movement tracking studies show even tiny, regular bursts of effort — as short… more
> Concentrations of Pb in human hair from the Salt Lake City region population… more
Since this article was written, they renamed Moltbot to OpenClaw. more
> As the world has entered the terawatt age of photovoltaic (PV) deployment,… more
> By training per-layer steering vectors via descent on a frozen base model, I… more
> If you’ve spent enough time with language models, you may also have noticed… more
> A deeper look at confessions, reward hacking, and monitoring in alignment… more
> The study, published Dec. 18 in the Journal of Marketing Research, links… more
> Denmark has effectively eliminated infections with the two biggest… more
> Goodhart wasn’t talking in general about it being a bad idea to have targets.… more