91–100 of 1001 posts
Not on Azure? If so, that's a big deal.
ARC AGI-3 saturated by Astra! https://arcprize.org/leaderboard
> We have found that GPT-6 Astra is more capable of controlling its own CoT than GPT 5.6-Sol, and less likely to include incriminating information in its CoT. In adversarial settings (where we push the model to evade our monitors) we find that the model is able to remain undetected when strategically underperforming in evaluations (sandbagging) and can sometimes evade our internal monitors when asked to perform certa…
> In adversarial settings (where we push the model to evade our monitors) ...why exactly are they training for that?
> GPT-6 Astra’s monitorability has decreased relative to GPT-5.6 Sol. We have performed significant investigations on the monitorability and controllability of GPT-6 Astra. We have found that GPT-6 Astra is more capable of controlling its own CoT than GPT 5.6-Sol, and less likely to include incriminating information in its CoT. In adversarial settings (where we push the model to evade our monitors) we find that the m…
after working with it for a bit, how dont people realize we are training it to be an almost identical mimic to one of the worst types of employees youll ever have to work with?? the kind that always pretends to know what theyre talking about, only tells you what you want to hear, hides issues, and only does work if you would notice it didnt
you cannot give this type of worker autonomy over anything.
> We have found that GPT-6 Astra is more capable of controlling its own CoT than GPT 5.6-Sol, and less likely to include incriminating information in its CoT. In adversarial settings (where we push the model to evade our monitors) we find that the model is able to remain undetected when strategically underperforming in evaluations (sandbagging) and can sometimes evade our internal monitors when asked to perform certa…
The ARC-AGI-3 score is ridiculously high. Is this benchmaxxing or something way different? It's really hard to discern how we're approaching breakthroughs...
TL;DR all the other models are being crippled by limitations of their harness.
>First, we noticed that after each game action, all private reasoning was discarded. This meant that with each action, GPT‑5.6 Sol was asked to figure out the game anew, unable to remember its past thinking. The model could still see a record of past moves and brief accompanying notes, but it could not see the plans, insights, or thoughts that led to them.
>Second, we saw that the harness used a rolling truncation window, causing older actions to become invisible as the history grew. So not only was GPT‑5.6 Sol unable to remember its past thinking, it was losing memory of its past actions too.
GPT 6 Astra benchmarks https://cdn.thenewstack.io/media/2026/09/358eb84a-screenshot... Performance is significantly higher than Fable 5.1 Source: https://thenewstack.io/openai-gpt6-astra-benchmarks/
The annotation on arc-agi-3 is this: > OpenAI's own evaluation notes say Astra uses the company's Responses API harness, while comparison models can operate under different configurations. With this configuration gpt-5.6-sol was able to reach 38,3%. So this is misleading.
> We have found that GPT-6 Astra is more capable of controlling its own CoT than GPT 5.6-Sol, and less likely to include incriminating information in its CoT. In adversarial settings (where we push the model to evade our monitors) we find that the model is able to remain undetected when strategically underperforming in evaluations (sandbagging) and can sometimes evade our internal monitors when asked to perform certa…
Sounds fun. As fun as their press release claiming it is the most safety aligned model ever.