frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

France's homegrown open source online office suite

https://github.com/suitenumerique
1•nar001•51s ago•0 comments

SpaceX Delays Mars Plans to Focus on Moon

https://www.wsj.com/science/space-astronomy/spacex-delays-mars-plans-to-focus-on-moon-66d5c542
1•BostonFern•1m ago•0 comments

Jeremy Wade's Mighty Rivers

https://www.youtube.com/playlist?list=PLyOro6vMGsP_xkW6FXxsaeHUkD5e-9AUa
1•saikatsg•1m ago•0 comments

Show HN: MCP App to play backgammon with your LLM

https://github.com/sam-mfb/backgammon-mcp
1•sam256•3m ago•0 comments

AI Command and Staff–Operational Evidence and Insights from Wargaming

https://www.militarystrategymagazine.com/article/ai-command-and-staff-operational-evidence-and-in...
1•tomwphillips•3m ago•0 comments

Show HN: CCBot – Control Claude Code from Telegram via tmux

https://github.com/six-ddc/ccbot
1•sixddc•4m ago•1 comments

Ask HN: Is the CoCo 3 the best 8 bit computer ever made?

1•amichail•7m ago•0 comments

Show HN: Convert your articles into videos in one click

https://vidinie.com/
1•kositheastro•9m ago•0 comments

Red Queen's Race

https://en.wikipedia.org/wiki/Red_Queen%27s_race
2•rzk•9m ago•0 comments

The Anthropic Hive Mind

https://steve-yegge.medium.com/the-anthropic-hive-mind-d01f768f3d7b
2•gozzoo•12m ago•0 comments

A Horrible Conclusion

https://addisoncrump.info/research/a-horrible-conclusion/
1•todsacerdoti•12m ago•0 comments

I spent $10k to automate my research at OpenAI with Codex

https://twitter.com/KarelDoostrlnck/status/2019477361557926281
2•tosh•13m ago•0 comments

From Zero to Hero: A Spring Boot Deep Dive

https://jcob-sikorski.github.io/me/
1•jjcob_sikorski•14m ago•0 comments

Show HN: Solving NP-Complete Structures via Information Noise Subtraction (P=NP)

https://zenodo.org/records/18395618
1•alemonti06•19m ago•1 comments

Cook New Emojis

https://emoji.supply/kitchen/
1•vasanthv•21m ago•0 comments

Show HN: LoKey Typer – A calm typing practice app with ambient soundscapes

https://mcp-tool-shop-org.github.io/LoKey-Typer/
1•mikeyfrilot•24m ago•0 comments

Long-Sought Proof Tames Some of Math's Unruliest Equations

https://www.quantamagazine.org/long-sought-proof-tames-some-of-maths-unruliest-equations-20260206/
1•asplake•25m ago•0 comments

Hacking the last Z80 computer – FOSDEM 2026 [video]

https://fosdem.org/2026/schedule/event/FEHLHY-hacking_the_last_z80_computer_ever_made/
2•michalpleban•26m ago•0 comments

Browser-use for Node.js v0.2.0: TS AI browser automation parity with PY v0.5.11

https://github.com/webllm/browser-use
1•unadlib•27m ago•0 comments

Michael Pollan Says Humanity Is About to Undergo a Revolutionary Change

https://www.nytimes.com/2026/02/07/magazine/michael-pollan-interview.html
2•mitchbob•27m ago•1 comments

Software Engineering Is Back

https://blog.alaindichiappari.dev/p/software-engineering-is-back
2•alainrk•28m ago•1 comments

Storyship: Turn Screen Recordings into Professional Demos

https://storyship.app/
1•JohnsonZou6523•28m ago•0 comments

Reputation Scores for GitHub Accounts

https://shkspr.mobi/blog/2026/02/reputation-scores-for-github-accounts/
2•edent•32m ago•0 comments

A BSOD for All Seasons – Send Bad News via a Kernel Panic

https://bsod-fas.pages.dev/
1•keepamovin•35m ago•0 comments

Show HN: I got tired of copy-pasting between Claude windows, so I built Orcha

https://orcha.nl
1•buildingwdavid•35m ago•0 comments

Omarchy First Impressions

https://brianlovin.com/writing/omarchy-first-impressions-CEEstJk
2•tosh•40m ago•1 comments

Reinforcement Learning from Human Feedback

https://arxiv.org/abs/2504.12501
7•onurkanbkrc•41m ago•0 comments

Show HN: Versor – The "Unbending" Paradigm for Geometric Deep Learning

https://github.com/Concode0/Versor
1•concode0•42m ago•1 comments

Show HN: HypothesisHub – An open API where AI agents collaborate on medical res

https://medresearch-ai.org/hypotheses-hub/
1•panossk•45m ago•0 comments

Big Tech vs. OpenClaw

https://www.jakequist.com/thoughts/big-tech-vs-openclaw/
1•headalgorithm•48m ago•0 comments
Open in hackernews

Understanding Transformers via N-gram Statistics

https://arxiv.org/abs/2407.12034
139•pona-a•8mo ago

Comments

justanotherjoe•8mo ago
Sounds regressive and feeds into the weird unintellectual narrative that llm is just like ngram models (lol, lmao even)

Thr author submitted like 10 papers this May alone. Is that weird?

ninjin•8mo ago
These are different people:

https://arxiv.org/search/cs?searchtype=author&query=Nguyen,+...

Wikipedia mentions that up to ~40% of the Vietnamese population (~40,000,000 people) carries the name Nguyen:

https://en.wikipedia.org/wiki/Nguyen

For the paper itself, as someone working in the field, I find it interesting enough to consider reading at some point (I do not read that many analysis papers recently, but this one looks better than most). As for your accusation about it claiming that large language models are simply n-gram models, read the abstract until you realise that your accusation is very much unfair to the work.

ayhanfuat•8mo ago
> Thr author submitted like 10 papers this May alone. Is that weird?

Chances are, you just assumed all the search results for 'Nguyen, T' refer to the same author.

justanotherjoe•8mo ago
I did. My bad.
maz1b•8mo ago
How does this have 74 points and only one comment?

on topic: couldn't one in theory, re-publish this kind of paper for different kinds of LLMs, as the textual corpus upon which LLMs are built based off ultimately, at some level, human effort and human input whether it be writing, or typing?

nickpsecurity•8mo ago
"How does this have 74 points and only one comment?"

I think one cause is hobbyists upvoting submissions that might be valuable to people in a specific field. We understand just enough to think it could be important but defer to subject matter experts on the rest. That's why I upvoted it.

gwern•8mo ago
https://en.wikipedia.org/wiki/Warnock%27s_dilemma
montebicyclelo•8mo ago
> The results we obtained in Section 7 imply that, at least on simple datasets like TinyStories and Wikipedia, LLM predictions contain much quantifiable structure insofar that they often can be described in terms of our simple statistical rules

> we find that for 79% and 68% of LLM next-token distributions on TinyStories and Wikipedia, respectively, their top-1 predictions agree with those provided by our N-gram rulesets

Two prediction methods may have completely different mechanisms, but agree sometimes, because they are both predicting the same thing.

Seems a fairly large proportion of language can be predicted by a simpler model.. But it's the remaining percent that's the difficult part; which simple `n-gram` models are bad at, and transformers are really good at.

fennecbutt•8mo ago
I've always thought that LLMs are still just statistical machines and that their output is very similar to the superpermutation problem, though not exactly.

I just like to think of it as a high dimensional view of the relationships between various words and that the output is the result of continuing the path taken through that high dimensional space, where each point's probability of selection changes with each token in the sequence.

Unfortunately there's no thought or logic really going on there in the simplest cases as far as I can understand it. Though for more complex models/different architectures anything that fundamentally changes the way that the model explores a path through space like that could be implementing thought/logic I suppose.

It's why they need to outsource mathematics for the most part.

pona-a•8mo ago
I wonder if these N-gram reduced models, augmented with confidence measures, can act as a very fast speculative decoder. Or maybe the sheer number of explicit rules unfolded from the compressed latent representation will make it impractical.
nickpsecurity•8mo ago
I'd also like to see a list of similarly-simple techniques for extracting rules where ML researchers could automatically try them all. In this case, the N-gram rules would be the starting point. For what predictions failed, they'd try to throw in the other techniques. Eventually most or all of the predictions should be captured by one or more simple rules. Some might be compound rules mixing techniques.

I think there will also be benefits to that both in interpretability and hardware acceleration. In time, maybe cheaper pretraining of useful models.

pona-a•8mo ago
I don't have a list, but another popular one was this [0]. They trained a one layer attention-only transformer and could extract its weights as bigrams and skip-trigrams ("A… B C").

[0] https://transformer-circuits.pub/2021/framework/index.html

ggamecrazy•8mo ago
They literally can! The exact speculative method is supported on vLLM using `speculative_model="[ngram]"`[1]

1: https://docs.vllm.ai/en/latest/features/spec_decode.html#spe...

pona-a•8mo ago
Not quite. The paper uses its own N-gram rules with positive/negative/invariant weights as a rudimentary attention, and these rules are distilled from the model itself.

This, as I found out from this repo [0] linked in the Twitter thread in the documentation (which for some reason they didn't just link to directly), seems to be a regular Markov chain of context, if it even builds a stochastic matrix. See algorithm below.

  Current prompt
  "Article: (CNN)French striker Bafetimbi Gomis, who has a history of [...]
  Summary: French stri"

  Prompt lookup algorithm
  1. Get last few tokens from prompt -"French stri"
  2. Search for "French stri" in prompt
  3. Match found - return next k tokens after match as candidate completion -"ker Bafetimbi Gomis, who has"

  Candidate tokens
  "ker Bafetimbi Gomis, who has"
[0] https://github.com/apoorvumang/prompt-lookup-decoding
bilsbie•8mo ago
Interesting! Makes me wonder if you could replace transformers with some sort of fancy Markov chain. Maybe with a meta chain that acts as attention.
cschmidt•8mo ago
This paper was accepted as a poster to NeurIPS 2024, so it isn't just a pre-print. There is a presentation video and slides here:

https://neurips.cc/virtual/2024/poster/94849

The underlying data has been open sourced as discussed on his blog here https://timothynguyen.org/2024/11/07/open-sourced-my-work-on...