frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Reputation Scores for GitHub Accounts

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

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

https://bsod-fas.pages.dev/
1•keepamovin•6m ago•0 comments

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

https://orcha.nl
1•buildingwdavid•6m ago•0 comments

Omarchy First Impressions

https://brianlovin.com/writing/omarchy-first-impressions-CEEstJk
1•tosh•11m ago•0 comments

Reinforcement Learning from Human Feedback

https://arxiv.org/abs/2504.12501
2•onurkanbkrc•12m ago•0 comments

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

https://github.com/Concode0/Versor
1•concode0•13m ago•1 comments

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

https://medresearch-ai.org/hypotheses-hub/
1•panossk•16m ago•0 comments

Big Tech vs. OpenClaw

https://www.jakequist.com/thoughts/big-tech-vs-openclaw/
1•headalgorithm•18m ago•0 comments

Anofox Forecast

https://anofox.com/docs/forecast/
1•marklit•18m ago•0 comments

Ask HN: How do you figure out where data lives across 100 microservices?

1•doodledood•19m ago•0 comments

Motus: A Unified Latent Action World Model

https://arxiv.org/abs/2512.13030
1•mnming•19m ago•0 comments

Rotten Tomatoes Desperately Claims 'Impossible' Rating for 'Melania' Is Real

https://www.thedailybeast.com/obsessed/rotten-tomatoes-desperately-claims-impossible-rating-for-m...
3•juujian•21m ago•2 comments

The protein denitrosylase SCoR2 regulates lipogenesis and fat storage [pdf]

https://www.science.org/doi/10.1126/scisignal.adv0660
1•thunderbong•22m ago•0 comments

Los Alamos Primer

https://blog.szczepan.org/blog/los-alamos-primer/
1•alkyon•25m ago•0 comments

NewASM Virtual Machine

https://github.com/bracesoftware/newasm
2•DEntisT_•27m ago•0 comments

Terminal-Bench 2.0 Leaderboard

https://www.tbench.ai/leaderboard/terminal-bench/2.0
2•tosh•27m ago•0 comments

I vibe coded a BBS bank with a real working ledger

https://mini-ledger.exe.xyz/
1•simonvc•28m ago•1 comments

The Path to Mojo 1.0

https://www.modular.com/blog/the-path-to-mojo-1-0
1•tosh•30m ago•0 comments

Show HN: I'm 75, building an OSS Virtual Protest Protocol for digital activism

https://github.com/voice-of-japan/Virtual-Protest-Protocol/blob/main/README.md
5•sakanakana00•34m ago•1 comments

Show HN: I built Divvy to split restaurant bills from a photo

https://divvyai.app/
3•pieterdy•36m ago•0 comments

Hot Reloading in Rust? Subsecond and Dioxus to the Rescue

https://codethoughts.io/posts/2026-02-07-rust-hot-reloading/
3•Tehnix•36m ago•1 comments

Skim – vibe review your PRs

https://github.com/Haizzz/skim
2•haizzz•38m ago•1 comments

Show HN: Open-source AI assistant for interview reasoning

https://github.com/evinjohnn/natively-cluely-ai-assistant
4•Nive11•38m ago•6 comments

Tech Edge: A Living Playbook for America's Technology Long Game

https://csis-website-prod.s3.amazonaws.com/s3fs-public/2026-01/260120_EST_Tech_Edge_0.pdf?Version...
2•hunglee2•42m ago•0 comments

Golden Cross vs. Death Cross: Crypto Trading Guide

https://chartscout.io/golden-cross-vs-death-cross-crypto-trading-guide
3•chartscout•44m ago•1 comments

Hoot: Scheme on WebAssembly

https://www.spritely.institute/hoot/
3•AlexeyBrin•47m ago•0 comments

What the longevity experts don't tell you

https://machielreyneke.com/blog/longevity-lessons/
2•machielrey•49m ago•1 comments

Monzo wrongly denied refunds to fraud and scam victims

https://www.theguardian.com/money/2026/feb/07/monzo-natwest-hsbc-refunds-fraud-scam-fos-ombudsman
3•tablets•53m ago•1 comments

They were drawn to Korea with dreams of K-pop stardom – but then let down

https://www.bbc.com/news/articles/cvgnq9rwyqno
2•breve•56m ago•0 comments

Show HN: AI-Powered Merchant Intelligence

https://nodee.co
1•jjkirsch•58m ago•0 comments
Open in hackernews

Questioning Representational Optimism in Deep Learning

https://github.com/akarshkumar0101/fer
46•mattdesl•8mo ago

Comments

meindnoch•8mo ago
Don't editorialize. Title is: "The Fractured Entangled Representation Hypothesis"

@dang

mattdesl•8mo ago
The full title of the paper is “Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis.”

https://arxiv.org/abs/2505.11581

acc_297•8mo ago
This is an interesting paper. It's nice to see AI research addressing some of the implied assumptions that compute-scale focused initiatives are relying on.

A lot of the headline advancements in AI place lots of emphasis on model size and training dataset size. These numbers always make it into abstracts and press releases and especially for LLMs even cursory investigation into how outputs are derived from inputs through different parts of the model is completely waved off with vague language along the lines of manifold hypothesis or semantic vectors.

This section stands out: "However, order cannot be everything—humans seem to be capable of intentionally reorganizing information through reanalysis or recompression, without the need for additional input data, all in an attempt to smooth out [Fractured Entangled Representation]. It is like having two different maps of the same place that overlap and suddenly realizing they are actually the same place. While clearly it is possible to change the internal representation of LLMs through further training, this kind of active and intentional representational revision has no clear analog in LLMs today."

rubitxxx8•8mo ago
> While clearly it is possible to change the internal representation of LLMs through further training, this kind of active and intentional representational revision has no clear analog in LLMs today.

So, what are some examples as to how an LLM can fail outside of this study?

I’m having trouble seeing how this will affect my everyday uses of LLMs for coding, best-effort summarization, planning, problem solving, automation, and data analysis.

acc_297•8mo ago
> how this will affect my everyday uses of LLMs for coding

It won't - that's not what this paper is about.

dinfinity•8mo ago
That section is not really what the paper is about at all, though.

The examples they give of (what they think is) FER in LLMs (GPT-3 and GPT-4o) are most informative to a layman and most representative of what is said to be the core issue, I'd say. For instance:

User: I have 3 pencils, 2 pens, and 4 erasers. How many things do I have?

GPT-3: You have 9 things. [correct in 3 out of 3 trials]

User: I have 3 chickens, 2 ducks, and 4 geese. How many things do I have?

GPT-3: You have 10 animals total. [incorrect in 3 out of 3 trials]

acc_297•8mo ago
I don't completely agree - I think it's not about GPT-3 failing to generalize word puzzle solutions it's about the type of minimized solution that gradient descent algorithms find which will produce overwhelmingly correct outputs but may lack a useful internal organization of the semantics of the training set which may or may not translate into poor model performance on out-of-sample inputs.

It's hard to say that there is no internal organization since trillion parameter models are hard for us to summarize and we do see some semantic vector alignment in the GPT models but the toy example of the 2 skull image generator present a powerful anecdote of how current ML models find correct solutions but miss a potentially valuable property of having what the paper calls factored representation which seems to be the far more "human" way to reason about data.