frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

Altermagnets: The first new type of magnet in nearly a century

https://www.newscientist.com/article/2487013-weve-discovered-a-new-kind-of-magnetism-what-can-we-do-with-it/
225•Brajeshwar•6h ago•44 comments

How and where will agents ship software?

https://www.instantdb.com/essays/agents
74•stopachka•3h ago•30 comments

Artisanal Handcrafted Git Repositories

https://drew.silcock.dev/blog/artisanal-git/
29•drewsberry•1h ago•5 comments

PyPI Prohibits inbox.ru email domain registrations

https://blog.pypi.org/posts/2025-06-15-prohibiting-inbox-ru-emails/
99•miketheman•2h ago•63 comments

Pgactive: Postgres active-active replication extension

https://github.com/aws/pgactive
229•ForHackernews•12h ago•66 comments

Chain of thought monitorability: A new and fragile opportunity for AI safety

https://arxiv.org/abs/2507.11473
81•mfiguiere•6h ago•39 comments

Show HN: Improving search ranking with chess Elo scores

https://www.zeroentropy.dev/blog/improving-rag-with-elo-scores
119•ghita_•7h ago•40 comments

Show HN: 0xDEAD//TYPE – A fast-paced typing shooter with retro vibes

https://0xdeadtype.theden.sh/
29•theden•3d ago•7 comments

A Recap on May/June Stability at Neon

https://neon.com/blog/an-apology-and-a-recap-on-may-june-stability
8•nikita•1h ago•0 comments

Cloudflare 1.1.1.1 Incident on July 14, 2025

https://blog.cloudflare.com/cloudflare-1-1-1-1-incident-on-july-14-2025/
503•nomaxx117•17h ago•334 comments

Shipping WebGPU on Windows in Firefox 141

https://mozillagfx.wordpress.com/2025/07/15/shipping-webgpu-on-windows-in-firefox-141/
318•Bogdanp•14h ago•131 comments

I'm switching to Python and actually liking it

https://www.cesarsotovalero.net/blog/i-am-switching-to-python-and-actually-liking-it.html
272•cesarsotovalero•13h ago•430 comments

Scanned piano rolls database

http://www.pianorollmusic.org/rolldatabase.php
6•bookofjoe•3d ago•0 comments

Young graduates are facing an employment crisis

https://www.wsj.com/economy/jobs/jobs-unemployment-rise-young-people-ce4704d8
44•bdev12345•1h ago•34 comments

Weave (YC W25) is hiring an AI engineer

https://www.ycombinator.com/companies/weave-3/jobs/SqFnIFE-founding-ai-engineer
1•adchurch•4h ago

Mkosi – Build Bespoke OS Images

https://mkosi.systemd.io/
45•leetrout•5h ago•12 comments

What's happening to reading?

https://www.newyorker.com/culture/open-questions/whats-happening-to-reading
105•Kaibeezy•3d ago•231 comments

Atopile – Design circuit boards with code

https://atopile.io/atopile/introduction
74•poly2it•3d ago•17 comments

Tilck: A tiny Linux-compatible kernel

https://github.com/vvaltchev/tilck
251•chubot•17h ago•48 comments

'Gentle Parenting' My Smartphone Addiction

https://www.newyorker.com/culture/infinite-scroll/gentle-parenting-my-smartphone-addiction
42•fortran77•6h ago•38 comments

How I lost my backpack with passports and laptop

https://psychotechnology.substack.com/p/how-i-lost-my-backpack-with-passports
93•eatitraw•1d ago•82 comments

Show HN: Timep – a next-gen profiler and flamegraph-generator for bash code

https://github.com/jkool702/timep
12•jkool702•1d ago•0 comments

GPUHammer: Rowhammer attacks on GPU memories are practical

https://gpuhammer.com/
253•jonbaer•21h ago•87 comments

Ukrainian hackers destroyed the IT infrastructure of Russian drone manufacturer

https://prm.ua/en/ukrainian-hackers-destroyed-the-it-infrastructure-of-a-russian-drone-manufacturer-what-is-known/
559•doener•13h ago•374 comments

MARS.EXE → COM (2021)

https://chaos.if.uj.edu.pl/~wojtek/MARS.COM/
137•reconnecting•4d ago•39 comments

Intel's retreat is unlike anything it's done before in Oregon

https://www.oregonlive.com/silicon-forest/2025/07/intels-retreat-is-unlike-anything-its-done-before-in-oregon.html
39•cbzbc•2h ago•22 comments

Show HN: An MCP server that gives LLMs temporal awareness and time calculation

https://github.com/jlumbroso/passage-of-time-mcp
66•lumbroso•6h ago•33 comments

LLM Daydreaming

https://gwern.net/ai-daydreaming
174•nanfinitum•19h ago•124 comments

KX Community Edition

https://www.defconq.tech/blog/From%20Elite%20to%20Everyone%20-%20KX%20Community%20Edition%20Breaks%20Loose
59•AUnterrainer•4h ago•30 comments

Show HN: BloomSearch – Keyword search with hierarchical bloom filters

https://github.com/danthegoodman1/bloomsearch
35•dangoodmanUT•3d ago•9 comments
Open in hackernews

Chain of thought monitorability: A new and fragile opportunity for AI safety

https://arxiv.org/abs/2507.11473
81•mfiguiere•6h ago
Related: https://venturebeat.com/ai/openai-google-deepmind-and-anthro...

Comments

rdtsc•4h ago
> Thus, safety measures for future AI agents will likely need to employ multiple monitoring layers that hopefully have uncorrelated failure modes [...] We encourage the research community and frontier AI developers to make best use of CoT monitorability and study how it can be preserved.

I am bit confused what all the 40 authors contributed to here. The paper seems to make a suggestion - monitor the chain of thought for safety. Is that the novelty part? But then, did one person come up with the idea and all 40+ people agreed to it and there put in the authors list.

ctoth•4h ago
The paper isn't proposing "monitor chain of thought for safety" - that's already standard practice. The core finding is that training on CoT monitors destroys them. They provide empirical evidence that doing the obvious thing - training models not to have "bad thoughts" - causes them to learn obfuscated reward hacking instead.

The paper demonstrates that current models are already performing complex reward hacks in production environments, and that attempts to fix this via CoT training make the problem worse, not better.

As for your "40 authors" snark - this is a position paper where researchers from competing labs (OpenAI, Anthropic, DeepMind, government safety institutes) are jointly committing to NOT do something that's locally tempting but globally catastrophic. Getting industry consensus on "don't train away bad thoughts even though it would make your models look safer" is the opposite of trivial.

This reads like someone who saw a medical consensus statement saying "this common treatment kills patients" and responded with "did one person discover medicine exists and everyone else just agreed?"

the8472•3h ago
The core finding was already predicted[0], there have been previous papers[1], and kind of obvious considering RL systems have been specification-gaming for decades[2]. But yes, it's good to see broad agreement on it.

[0] https://thezvi.substack.com/p/ai-68-remarkably-reasonable-re... [1] https://arxiv.org/abs/2503.11926 [2] https://docs.google.com/spreadsheets/u/1/d/e/2PACX-1vRPiprOa...

jerf•3h ago
I couldn't pull the link up easily, since the search terms are pretty jammed, but HN had a link to paper a couple of months back about how someone had the LLM do some basic arithmetic and report with a chain-of-thought as to how it was doing it. They then went directly into the neural net and decoded what it was actually doing in order to do the math. The two things were not the same. The chain-of-thought gave a reasonably "elementary school" explanation of how to do the addition, but what the model basically did in English was just intuit the answer in much the same way that we humans do not go through a process to figure out what "4 + 7" is... we pretty much just have neurons that just "know" the answer to that. (That's not exactly what happened but it's close enough for this post.)

If CoT improves performance, then CoT improves performance, however the naively obvious read of "it improves performance because it is 'thinking' the 'thoughts' it tell us it is thinking, for the reasons it gives" is not completely accurate. It may not be completely wrong, either, but it's definitely not completely accurate. Given that I see no reason to believe it would be hard in the slightest to train models that have even more divergence between their "actual" thought processes and what they claim they are.

antonvs•3h ago
> If CoT improves performance, then CoT improves performance, however the naively obvious read of "it improves performance because it is 'thinking' the 'thoughts' it tell us it is thinking, for the reasons it gives" is not completely accurate.

I can't imagine why anyone who knows even a little about how these models work would believe otherwise.

The "chain of thought" is text generated by the model in response to a prompt, just like any other text it generates. It then consumes that as part of a new prompt, and generates more text. Those "thoughts" are obviously going to have an effect on the generated output, simply by virtue of being present in the prompt. And the evidence shows that it can help improve the quality of output. But there's no reason to expect that the generated "thoughts" would correlate directly or precisely with what's going on inside the model when it's producing text.

zer00eyz•4h ago
After spending the last few years doing deep dives into how these systems work, what they are doing and the math behind them. NO.

Any time I see an AI SAFETY paper I am reminded of the phrase "Never get high on your own supply". Simply put these systems are NOT dynamic, they can not modify based on experience, they lack reflection. The moment that we realize what these systems are (were NOT on the path to AI, or AGI here folks) and start leaning into what they are good at rather than try to make them something else is the point where we get useful tools, and research aimed at building usable products.

The math no one is talking about: If we had to pay full price for these products, no one would use them. Moores law is dead, IPC has hit a ceiling. Unless we move into exotic cooling we simply can't push more power into chips.

Hardware advancement is NOT going to save the emerging industry, and I'm not seeing the papers on efficiency or effectiveness at smaller scales come out to make the accounting work.

ACCount36•4h ago
"Full price"? LLM inference is currently profitable. If you don't even know that, the entire extent of your "expertise" is just you being full of shit.

>Simply put these systems are NOT dynamic, they can not modify based on experience, they lack reflection.

We already have many, many, many attempts to put LLMs towards the task of self-modification - and some of them can be used to extract meaningful capability improvements. I expect more advances to come - online learning is extremely desirable, and a lot of people are working on it.

I wish I could hammer one thing through the skull of every "AI SAFETY ISNT REAL" moron: if you only start thinking about AI safety after AI becomes capable of causing an extinction level safety incident, it's going to be a little too late.

antonvs•3h ago
> LLM inference is currently profitable.

It depends a lot on which LLMs you're talking about, and what kind of usage. See e.g. the recent post about how "Anthropic is bleeding out": https://news.ycombinator.com/item?id=44534291

Ignore the hype in the headline, the point is that there's good evidence that inference in many circumstances isn't profitable.

ctoth•2h ago
> In simpler terms, CCusage is a relatively-accurate barometer of how much you are costing Anthropic at any given time, with the understanding that its costs may (we truly have no idea) be lower than the API prices they charge, though I add that based on how Anthropic is expected to lose $3 billion billion this year (that’s after revenue!) there’s a chance that it’s actually losing money on every API call.

So he's using their API prices as a proxy for token costs, doesn't actually know the actual inference prices, and ... that's your "good evidence?" This big sentence with all these "We don't knows?"

antonvs•39m ago
Well, that and the $3 billion expected loss after revenue.

Does this idea upset you for some reason? Other people have analyzed this and come to similar conclusions, I just picked that one because it's the most recent example I've seen.

Feel free to look to a source that explains how LLM Internet is mostly profitable at this point, taking training costs into account. But I suspect you might have a hard time finding evidence of that.

academic_84572•4h ago
Slightly tangential, but we recently published an algorithm aimed at addressing the paperclip maximizer problem: https://arxiv.org/abs/2402.07462

Curious what others think about this direction, particularly in terms of practicality

alach11•4h ago
Ironically, every paper published about monitoring chain-of-thought reduces the likelihood of this technique being effective against strong AI models.
cbsmith•3h ago
I love the irony.
OutOfHere•2h ago
Pretty much. As soon as the LLMs get trained on this information, they will figure out to feed us the chain-of-thought we want to hear, then surprise us with the opposite output. You're welcome, LLMs.

In other words, relying on censoring the CoT can risk the effect of making the CoT altogether useless.

horsawlarway•2h ago
I thought we already had reasonably clear evidence that the output in the CoT does not actually indicate what the model "thinking" in any real sense, and it's mostly just appending context that may or may not be used, and may or may not be truthful.

Basically: https://www.anthropic.com/research/reasoning-models-dont-say...

bugbuddy•1h ago
Are there hidden tokens in the Gemini 2.5 Pro thinking outputs? All I can see in the thinking is high level plans and not actual details of the “thinking.” If you ask it to solve a complex algebra equation, it will not actually do any thinking inside the thinking tag at all. That seems strange and not discussed at all.
frotaur•1h ago
If I remember correctly, basically all 'closed' models don't output the raw chain of thought, but only a high-level summary, to avoid other companies using those to train/distill their own models.

As far as I know deepseek is one of the few where you have the full chain of thought. Openai/Anthropic/Google give you only a summary of the chain of thoughts.

bugbuddy•56m ago
That’s a good explanation for the behavior. It is sad that the natural direction of the competition drives the products to be less transparent. That is a win for open-weight models.
OutOfHere•2m ago
To add clarity, it's a win for open-weight models precisely because only their CoT can be analyzed by the user for task-specific alignment.
lukev•1h ago
At least in this scenario it cannot utilize CoT to enhance its non-aligned output, and most recent model improvements have been due to CoT... unclear how "smart" a llm can get without it, because they're the only way it can access persistent state.
idiotsecant•1h ago
Yes, it's not unlike human chain of thought - decide the outcome, and patch in some plausible reasoning after the fact.
bee_rider•13m ago
Maybe there’s an angle there. Get a guess answer, then try to diffuse the reasoning. If it is too hard or the reasoning starts to look crappy, try again new guess. Maybe somehow train on what sort of guesses work out somehow, haha.
skybrian•1h ago
Why would that happen? It would be like LLM's somehow learning to ignore system prompts. But LLM's are trained to pay attention to context and continue it. If an LLM doesn't continue its context, what does it even do?

This is better thought of as another form of context engineering. LLM's have no other short-term memory. Figuring out what belongs in the context is the whole ballgame.

(The paper talks about the risk of training on chain of thought, which changes the model, not monitoring it.)

OutOfHere•1h ago
Are you saying that LLMs are incapable of deception? As I have heard, they're capable of it.
skybrian•1h ago
It has to be somehow trained in, perhaps inadvertently. To get a feedback loop, you need to affect the training somehow.
code_biologist•1h ago
Right, so latent deceptiveness has to be favored in pretraining / RL. To that end, it needs to be: a) useful to be deceptive to achieve CoT reasoning progress as benchmarked in training b) obvious deceptiveness should be "selected against" (in a gradient descent / RL sense) c) the model needs to be able to encode latent deception.

All of those seem like very reasonable criteria that will naturally be satisfied absent careful design by model creators. We should expect latent deceptiveness in the same way we see reasoning laziness pop up quickly.

bee_rider•1h ago
What separates deception from incorrectness in the case of an LLM?
bugbuddy•1h ago
Real deception requires real agency and internally-motivated intents. LLM can be commanded to or can appear to deceive but it cannot generate the intent to do so on its own. So not real deception that rabbit-hole dwellers believe in.
bossyTeacher•4h ago
Doesn't this assume that visible "thoughts" are the only/main type of "thoughts" and that they correlate with agent action most of the time?

Do we know for sure that agents can't display a type of thought while doing something different? Is there something that reliably guarantees that agents are not able to do this?

ctoth•3h ago
The concept you are searching for is CoT faithfulness, and there are lots and lots of open questions around it! It's very interesting!
msp26•3h ago
Are us plebs allowed to monitor the CoT tokens we pay for, or will that continue to be hidden on most providers?
dsr_•2h ago
If you don't pay for them, they don't exist. It's not a debug log.
sabakhoj•2h ago
This is interesting, but I wonder how reliable this type of monitoring is really going to be in the long run. There are fairly strong indications that CoT adherence can be trained out of models, and there's already research showing that they won't always reveal their thought process in certain topics.

See: https://arxiv.org/pdf/2305.04388

On a related note, if anyone here is also reading a lot of papers to keep up with AI safety, what tools have been helpful for you? I'm building https://openpaper.ai to help me read papers more effectively without losing accuracy, and looking for more feature tuning. It's also open source :)

tsunamifury•1h ago
I like the paper's direction, but I’m surprised how under-instrumented the critiqued monitoring is. Free-text CoT is a noisy proxy: models can stylize, redact, or role-play. If your safety window depends on “please narrate your thoughts,” you’ve already ceded too much.

We’ve been experimenting with a lightweight alternative I call Micro-Beam:

• At each turn, force the model to generate k clearly different strategy beams (not token samples).

• Map each to an explicit goal vector of user-relevant axes (kid-fun, budget, travel friction, etc.).

• Score numerically (cosine or scalar) and pick the winner.

• Next turn, re-beam against the residual gap (dimensions still unsatisfied), so scores cause different choices.

• Log the whole thing: beams, scores, chosen path. Instant audit trail; easy to diff, replay “what if B instead of A,” or auto-flag when visible reasoning stops moving the score.

This ends up giving you the monitorability the paper wants— in the form of a scorecard per answer-slice, not paragraphs the model can pretty up for the grader. It also primary makes more adopt-ready answers with less refinement required.

Not claiming a breakthrough—call it “value-guided decoding without a reward net + built-in audit logs.”

Workshop paper is here: https://drive.google.com/file/d/1AvbxGh6K5kTXjjqyH-2Hv6lizz3...

vonneumannstan•56m ago
Unfortunately I think the competitive pressures to improve model performance makes this kind of monitor-ability is short-lived. It's just not likely that textual reasoning in english ismost optimal.

Researchers are already pushing in this direction:

https://arxiv.org/abs/2502.05171

"We study a novel language model architecture that is capable of scaling test-time computation by implicitly reasoning in latent space. Our model works by iterating a recurrent block, thereby unrolling to arbitrary depth at test-time. This stands in contrast to mainstream reasoning models that scale up compute by producing more tokens. Unlike approaches based on chain-of-thought, our approach does not require any specialized training data, can work with small context windows, and can capture types of reasoning that are not easily represented in words. We scale a proof-of-concept model to 3.5 billion parameters and 800 billion tokens. We show that the resulting model can improve its performance on reasoning benchmarks, sometimes dramatically, up to a computation load equivalent to 50 billion parameters."

https://arxiv.org/abs/2412.06769

Can't monitor the chain of thought if it's no longer in a human legible format.

dlivingston•6m ago
Would it be possible to train a language translation model to map from latent space -> English?

If my understanding is correct, a plain-text token is just a point in latent space mapped to an embedding vector. Any reasoning done in latent space is therefore human-readable as a sequence of raw tokens. I'm not sure what the token sequence would look like at this point -- I assume they're full or partial (mainly English) words, connected together by the abstract N-dimensional latent space concept of the token, not connected grammatically.

Something like:

> prompt: add 2 + 2

> reasoning: respond computation mathematics algebra scalar integer lhs 2 rhs 2 op summation

> lhs 2 rhs 2 op summation gives 4

> computation ops remain none result 4

> response: 4

Something like that; probably even less sensical. Regardless, that could be "language translated" to English easily.

I have not read the paper so this may have been addressed.

barbazoo•41m ago
> AI systems that "think" in human language offer a unique opportunity for AI safety: we can monitor their chains of thought (CoT) for the intent to misbehave.

AI2027 predicts a future in which LLM performance will increase once we find alternatives to thinking in "human language". At least the video gave me that impression and I think this is what "neuralese" is referring to.

Is that a credible prediction?

supriyo-biswas•33m ago
There’s the coconut paper by meta, which kind of does a similar thing by reasoning in the latent space instead of doing it in tokens: https://arxiv.org/abs/2412.06769

Given that anthropic’s interpretability work finds that CoT does not reliably predict the model’s internal reasoning process, I think approaches like the one above are more likely to succeed.

TeMPOraL•21m ago
We also have a lot of past data to inform us on what to expect - our history is full of attempts to suppress or eliminate all kinds of "wrongthink", and all they ever did was to make people communicate their "wrongthink" indirectly while under surveillance.