标签:#News
找到 9773 篇相关文章
How p-hacking built the world's most expensive safety regime
TabFM: A zero-shot foundation model for tabular data
Ask HN: Since when does Craigslist's front page have emojis?
Today I noticed the inclusion of emojis in Craigslist's listings/categories: https://www.craigslist.org/area/sfbay . Now, Craigslist, as a legacy of the 1990s web, has for a long time stubbornly maintained its minimalist style, to the point where several "modern" startups have popped up to try and offer Craigslist-like services to new generations. So why this change? And what's with the timing? It's coinciding with the wanton proliferation of emojis everywhere courtesy of everyone's favorite GPT
Gone but Not Forgotten: Recovering the Dead Web
Ship traces journey Spanish Armada sailors made in 1588
US Army Women Are More Likely to Be Killed by Army Men Than by War
Understanding lattice risks: Many differences between marketing and reality
Meta's brain-scanning system reads sentences non-invasively, code open source
Superpowers 6
Breaking the Bird Barrier: Scientist Decodes Zebra Finch Language
Amazon fined $2.25 million for failing to help identity theft victims
The Federal Trade Commission fined Amazon $2.25 million to settle claims that the company failed to help customers who fell victim to identity theft, as reported earlier by Bloomberg. In its complaint, the FTC accuses Amazon of refusing to provide customers with information about purchases made with fraudulent accounts, in violation of the Fair Credit […]
Daisy Sound’s First Headphones Are Premium, High-Quality—and Just a Little Bit Cheaper
The California-based startup enters a crowded headphone market with a $399 chiseled-aluminum gamble.
Claude Code Just Got 5x More Expensive
Local Reasoning for Global Properties
Reddit Age Verification in Europe
Show HN: Morph Reflexes – Multi-head classifiers for agent traces
The most common failures for production agents are behavioral: looping, reasoning leakage, user frustration, and more. Using a frontier model like GPT or Sonnet to judge every turn is too expensive and slow to run at scale. To solve this, we built Reflexes: semantic signals from agent traces, served fast and cheap over API. Built on custom kernels and a custom inference engine forked from vLLM. Under the hood, it is a small LLM architected around multi-head inference. Small models need to be tra