GPT-5.6 • Hacker News Board

Re: GPT-5.6

#691

Post by irthomasthomas » Thu, Jul 09, 2026, 11:19 PM UTC

astrange wrote: post #471

adam_arthur wrote: post #267

Earlier quoted context omitted.

Information density of the prompt is the most important factor in my experience. And interestingly, LLMs seem particularly bad at writing prompts for other LLMs for this reason (you can guide them to be more dense, just speaking by default). Conciseness is usually a byproduct of information density though.

> LLMs seem particularly bad at writing prompts for other LLMs for this reason Claude is terrible at this! Probably for the same reason that its writing style in prose is so annoying and full of claudisms.

Claude use to be leader, too. Their metaprompt was great at the time with opus 3

Re: GPT-5.6

#692

Post by matheusmoreira » Thu, Jul 09, 2026, 11:20 PM UTC

postalcoder wrote: post #27

Syntaf wrote: post #6

Ok long time Claude Code user here; lately I've started to realize there's other great models out there I should be trying, but I'm hesitant to leave Claude Code behind for something new. What's the consensus today on codex vs claude code, does it really matter anymore?

There is so much less drama involved with the Codex world. You don't realize how oppressive CC is until you've escaped it. Outages, weird restrictions, degradation, accelerated usage, etc etc etc.

> accelerated usage

Can you post more information about this?

Re: GPT-5.6

#693

Post by matheusmoreira » Thu, Jul 09, 2026, 11:21 PM UTC

wwind123 wrote: post #127

Syntaf wrote: post #6

Ok long time Claude Code user here; lately I've started to realize there's other great models out there I should be trying, but I'm hesitant to leave Claude Code behind for something new. What's the consensus today on codex vs claude code, does it really matter anymore?

I've been using Claude Code, Codex, Gemini (now Antigravity) at the same time for half year now, ever since I dipped my toe into agentic coding. I'd say in general Claude Code and Codex are equally powerful, Gemini is lagging behind. One thing I appreciate with Codex is, OpenAI nowadays sometimes just gives you quota resets you can bank, so when you use up weekly quota before the week ends, you could just reset the q…

> OpenAI nowadays sometimes just gives you quota resets you can bank

That's actually pretty awesome. Anthropic's random resets often have me scrambling to launch huge sessions to make the most of them before the weekly rollover. The gacha-like mechanics are maddening.

Re: GPT-5.6

#694

Post by RagnarD » Thu, Jul 09, 2026, 11:23 PM UTC

Annoyingly, the new ChatGPT app which folds in Codex, no longer recognizes Shift-Tab to toggle plan mode. Irritatingly you have to enter /plan. OpenAI, fix this!

Re: GPT-5.6

#695

Post by bansimonw » Thu, Jul 09, 2026, 11:23 PM UTC

simianwords wrote: post #218

meetpateltech wrote: post #154

GPT-5.6 Sol sets a new SOTA on ARC-AGI-3: 7.8% Sol is the first verified frontier model to ever beat an ARC-AGI-3 game https://arcprize.org/results/openai-gpt-5-6

Very interesting. My prediction is that Mythos would outperform Sol. Also what does this tell about Yann LeCuns whole world model theory? Bro has been going on and on about it. He has made multiple wrong predictions on the trajectory of LLMs. At some point his claim should be fully falsified no?

[dead]

Re: GPT-5.6

#696

Post by marsven_422 » Thu, Jul 09, 2026, 11:25 PM UTC

anticorporate wrote: post #141

ravenstine wrote: post #94

Earlier quoted context omitted.

> Avoid generic brevity instructions That part is confusing because it's not like they provide an example of how default GPT-5.6 output compares with GPT-5.5 both with default output and prompted for brevity. Whenever I use such prompts, it's usually because I want the model to give me the gist in a few sentences. I'd be stunned if GPT-5.6 was that concise by default. I would think that could "break" a lot of things…

It seems like the way brevity instructions have changed is mis-aligned with how most people would expect to use them or are currently using them. Here's the example they give: > Instead of asking for the shortest possible answer, replace brevity instructions with prioritization: > Lead with the conclusion. Include the evidence needed to support it, any material caveat, and the next action. Omit secondary detail and r…

[dead]

Re: GPT-5.6

#697

Post by erichocean » Thu, Jul 09, 2026, 11:32 PM UTC

epolanski wrote: post #620

erichocean wrote: post #164

Earlier quoted context omitted.

I use Claude for planning, writing CRs, and code review. Codex writes all of the code, no exceptions. Works great, especially when you ask Claude to break up large CRs into roughly 10 minutes of Codex work each.

I do exactly the opposite.

I think the key is to get two LLMs looking at the same problem.

I use Codex because it's better at the kind of code I need written (math-heavy, 3D geometry code).

But if I was doing mainly UI code, I would do the opposite.

Re: GPT-5.6

#698

Post by robertwt7 » Thu, Jul 09, 2026, 11:32 PM UTC

it seems like 5.6 SOL is better at almost everything than Mythos except Coding Benchmarks (except TerminalBench)? anyone knows why Mythos scores so high on SWEBench are they cheating or are they just optimised better for coding?

Re: GPT-5.6

#699

Post by mrandish » Thu, Jul 09, 2026, 11:34 PM UTC

derefr wrote: post #303

anticorporate wrote: post #141

Earlier quoted context omitted.

It seems like the way brevity instructions have changed is mis-aligned with how most people would expect to use them or are currently using them. Here's the example they give: > Instead of asking for the shortest possible answer, replace brevity instructions with prioritization: > Lead with the conclusion. Include the evidence needed to support it, any material caveat, and the next action. Omit secondary detail and r…

> Lead with conclusion. I would presume (perhaps falsely?) that an instruction like this would lead to the model presenting a conclusion not supported by the evidence, and potentially backtracking as it then tries to justify said conclusion. Yes, if deliberation happens, the model should figure out what it wants to say during that phase; but if you're using auto mode, the model is not going to be doing any deliberati…

Over hours of experimentation with various LLMs, I've found virtually any system prompt can cause unintended skewing of the model's output. Even just 5 to 8 short, direct words about length, tone or formatting can cause subtle yet significant changes in model output.

Longer, more detailed or conditional prompts always introduce an additional cognitive load as it checks every token it generates against the conditions. Making instructions more absolute (like: "Never do...") can increase the duration of compliance but at the cost of creating a significant center of attentional gravity. This can cause far more output distortion as the model devotes increasing portions of its attention budget to ensure compliance with a heavyweight requirement or prohibition. Every word in a global prompt is a trade-off between attention, compliance, drift, etc.

As someone used to thinking of computers as natural deterministic rule-followers, it's weird having to carefully wordsmith and A/B test even the simplest global prompts. It feels like coaxing a hyper-literal, emotionally sensitive, spectrum-ish toddler to comply but without being so strict it gets 'upset' or spirals into hyper-focusing.

Re: GPT-5.6

#700

Post by DetroitThrow » Thu, Jul 09, 2026, 11:35 PM UTC

gorgmah wrote: post #518

Anyone else noticed the "Extended: Fable 5 is included in your weekly limit through July 12 blablabla" disappeared from claude code? Did they panic-delete the july 12th deadline ?

Looks like they reset everyone's Fable usage.