Feed aggregator
Ex-JPMorgan banker Chirayu Rana files fed civil rights suit; racism/sexual abuse
Manticore Search 28.4.4:Faster KNN, conversational search, installs and faceting
VibeUE: Unreal Engine Vibe Coding Tool
Article URL: https://github.com/kevinpbuckley/VibeUE
Comments URL: https://news.ycombinator.com/item?id=49079130
Points: 1
# Comments: 0
Newspeak
Article URL: https://github.com/dv-hart/newspeak
Comments URL: https://news.ycombinator.com/item?id=49079080
Points: 1
# Comments: 0
Show HN: I taught Claude Code to file Indian income-tax returns (ITR)
Article URL: https://github.com/karanb192/itr-wala
Comments URL: https://news.ycombinator.com/item?id=49079060
Points: 1
# Comments: 0
Storj begins court-supervised restructuring, proposes equity for token holders
Architecting Persistent Memory for Multi-Agent System [video]
Article URL: https://www.youtube.com/watch?v=lw52w7Pw86M\
Comments URL: https://news.ycombinator.com/item?id=49078958
Points: 2
# Comments: 0
Playing Computer Music via Am Radio on Computers Like Altair, IMSAI, PDP8, etc.
Article URL: https://github.com/rcl9/Playing-Computer-Music-via-AM-Radio-on-IMSAI-8080-PDP8-Sorcerer-TRS80-etc-in-the-1970s
Comments URL: https://news.ycombinator.com/item?id=49078937
Points: 1
# Comments: 1
Show HN: A macOS Port of Ventoy2disk
Article URL: https://github.com/fcjr/ventoy-mac
Comments URL: https://news.ycombinator.com/item?id=49078916
Points: 1
# Comments: 0
Sam Altman: "Never a Better Time to Do a Startup" [video]
Article URL: https://www.youtube.com/watch?v=ZIaOBAjvc38
Comments URL: https://news.ycombinator.com/item?id=49078904
Points: 1
# Comments: 0
39,535+ hours of open robot data
Article URL: https://datasets.bot/
Comments URL: https://news.ycombinator.com/item?id=49078866
Points: 1
# Comments: 0
Walk on Decomposed Subdomains
Article URL: https://clementjambon.github.io/wods/index.html#blogpost
Comments URL: https://news.ycombinator.com/item?id=49078856
Points: 1
# Comments: 0
ICE launching operation in New York City
Article URL: https://thehill.com/policy/national-security/5992642-ice-queens-long-island-raids/
Comments URL: https://news.ycombinator.com/item?id=49078844
Points: 3
# Comments: 0
Key Considerations for On-Premises Project Management Platform Planning
Article URL: https://ones.com/blog/10-key-considerations-for-on-premises-project-management-platform-planning/
Comments URL: https://news.ycombinator.com/item?id=49078839
Points: 1
# Comments: 0
Tracked Capabilities for Safer Agents
Article URL: https://martinodersky.substack.com/p/tracked-capabilities-for-safer-agents
Comments URL: https://news.ycombinator.com/item?id=49078826
Points: 2
# Comments: 0
Street videos 'exploiting' vulnerable people
Article URL: https://www.bbc.com/news/articles/c0rdglq4nvlo
Comments URL: https://news.ycombinator.com/item?id=49078804
Points: 1
# Comments: 0
For Some, So-Called ‘Skynet Day’ Came too Close to Sci-Fi After a Rogue Agent Hacked Into a Startup
Decades after it appeared in “The Terminator,” Skynet looks more like a forecast of the cyber incident in which a rogue AI system hacked into another AI company on its own.
The post For Some, So-Called ‘Skynet Day’ Came too Close to Sci-Fi After a Rogue Agent Hacked Into a Startup appeared first on SecurityWeek.
Show HN: Gemma 4 26B A4B running on an iPhone 17 Pro via model paging
I wanted to share a use case for Noema Overfit, a model-paging system available today in the Noema app. For disclosure, I founded Noema and I am part of the team that has helped develop the system.
The way it works is that the non-expert weights remain resident in memory, while the model’s routed expert weights are read from storage as needed. This makes it possible to run models that would otherwise exceed the device’s available memory, with the expected tradeoff of higher time to first token and slower generation.
For a 699-token initial prompt, I measured:
Prefill speed: 34.4 tokens/s Prefill time: 20.34 seconds Decode speed: 3.5 tokens/s
Photos of Overfit in action: https://noemaai.com/overfit/gemma4-iphone
The complete answer took approximately six minutes to generate, but the answer was correct. This is not intended for interactive, low-latency chat. The more interesting use case is allowing a constrained device to run a substantially more capable model when answer quality matters more than response time.
The same paging system can also be useful on lower-memory MacBooks that cannot ordinarily keep the complete model in unified memory.
I would be interested to hear where people think this tradeoff might be useful, and whether there are other models or workloads we should test.
Technical details: https://noemaai.com/overfit
Paged models: https://huggingface.co/NoemaAI-labs/Noema-Overfit
Comments URL: https://news.ycombinator.com/item?id=49078353
Points: 1
# Comments: 0
China Begins Making Homegrown DUV Lithography Chipmaking Tools
Trusty Boot Key, a Ventoy Alternative
Article URL: https://codeberg.org/aol/trusty-boot-key
Comments URL: https://news.ycombinator.com/item?id=49078332
Points: 1
# Comments: 0
