←
nano3,404 posts
nano avatar

nano

@nanulled

longtermism

◷ Joined October 2019▧ 808 Photos and videos
Artificial Analysis avatar
⟳ nano retweeted
Artificial Analysis@ArtificialAnlys·Jun 17

Z ai’s GLM-5.2 is the new leading open weights model on the Artificial Analysis Intelligence Index scoring 51 and it sits on the Pareto frontier of Intelligence vs Cost per Task @Zai_org’s GLM-5.2 is the same size as GLM-5.1 (744B total / 40B active parameters) but scores 11 points higher on the Intelligence Index v4.1, placing ahead of MiniMax-M3 (44) and DeepSeek V4 Pro (max, 44). On the first-party API it is priced in line with GLM-5.1 at $1.4/$4.4/$0.26 per 1M input/output/cache hit tokens Key results: ➤ GLM-5.2 is the leading open weights model on the Intelligence Index v4.1. At 51, it leads MiniMax-M3 (44), DeepSeek V4 Pro (max, 44) and Kimi K2.6 (43) ➤ Improvements across most evaluations, particularly scientific reasoning: GLM-5.2 gains over GLM-5.1 on most evaluations, led by scientific reasoning on CritPt (+16 points to 21%) and HLE (+12 points to 40%), alongside AA-LCR (+9 points to 71%), tau3 banking (+15 points to 27%) and SciCode (+7 points to 50%). TerminalBench v2.1 also improves (+16 points to 78%) and GPQA Diamond gains 3 points to 89% ➤ Leading open weights model on GDPval-AA v2 and competitive with proprietary models: GLM-5.2 scores 1524 on GDPval-AA v2, ahead of MiniMax-M3 (1418) and DeepSeek V4 Pro (max, 1328). This impressive result places GLM-5.2 in-line with proprietary models including GPT-5.5 (xhigh reasoning). GDPval-AA v2 builds on the original GDPval-AA by baselining Elo to human performance at 1000, introducing a rotating panel of frontier-model judges, and raising the turn limit from 100 to 250 for longer-horizon agent trajectories ➤ GLM-5.2 uses more output tokens per task than other leading open weights models: the model uses 43k output tokens per Intelligence Index task, up from GLM-5.1 (26k) and above MiniMax-M3 (24k), Kimi K2.6 (35k) and DeepSeek V4 Pro (max, 37k) ➤ On the Intelligence vs. Cost per Task Pareto Frontier: GLM-5.2 is on the Pareto frontier of the Intelligence vs Cost per Task chart, with the lowest cost per task among models at its intelligence level. GLM-5.2 costs ~$0.46 per task, compared to GLM-5.1 ($0.25), Kimi K2.6 ($0.31), MiniMax-M3 ($0.18) and DeepSeek V4 Pro (max, $0.05) Additional Model Details: ➤ License: MIT ➤ Size: 744B total parameters, 40B active parameters, equivalent to GLM-5.1 ➤ Context window: 1M tokens, up from 200K on GLM-5.1 ➤ Pricing: $1.4/$0.26/$4.4 per 1M input/cache hit/output tokens ➤ Availability: Alongside Z ai's first-party API, GLM-5.2 is available across third-party providers including @DeepInfra, @novita_labs, @nebiusai, @parasailnetwork , @SiliconFlowAI , @gmi_cloud , @Baseten and @FireworksAI_HQ

Post attachment
Z.ai avatar
⟳ nano retweeted
Z.ai@Zai_org·Jun 16

Introducing GLM-5.2: Frontier Intelligence, Open Weights - Significant improvements in coding and agentic tasks - Strong long-horizon capabilities with a 1M context window - Two levels of reasoning effort: GLM-5.2 (max) pushes the limits, while GLM-5.2 (high) strikes a strong balance between performance and token efficiency - MIT-licensed open weights - Same API pricing as GLM-5.1 Tech Blog: z.ai/blog/glm-5.2 Weights: huggingface.co/zai-org/GLM-5… API: docs.z.ai/guides/llm/glm-5.2 Coding Plan: z.ai/subscribe Chat: chat.z.ai

Post attachment
Arthur Mensch avatar
⟳ nano retweeted
Arthur Mensch@arthurmensch·Jun 16

We somehow got put in the spotlight the last few days! First we'd like to thank the organizers of the AI show for that, we can't get enough of this stuff. I'll say a few things about where we are and what we do.

Brian Fioca avatar
⟳ nano retweeted
Brian Fioca@bfioca·Jun 10

In light of Anthropic’s policy decision, I am withdrawing my amicus brief signature. I can’t truthfully argue they’re not a supply chain risk. 😞

I’ll be honest that it would have been much more difficult to defend Anthropic against the DoW incursion had that incident occurred after this one. This is the company literally telling their customers, “we reserve the right to silently sabotage you.” I’d still have defended them, because the government trying to destroy a firm is still wrong, but man would it have been a harder case to make.

spor avatar
⟳ nano retweeted
spor@sporadica·Jun 10

if I'm Dario, the reasoning for an IPO is to get flush with liquidity and then pay for the most insane lobbying effort the US has ever seen i'm calling it now, come back to this in 6 months

Post attachment

Today I'm publishing a new essay, Policy on the AI Exponential. AI is progressing extremely fast—much faster than the policy process was built to handle. The essay lays out where I think the technology is now, and the action needed to close the gap: darioamodei.com/post/policy-…

renji avatar
⟳ nano retweeted
renji@brickroad7·Jun 10

Fable is an early look into the bureaucratic hell Anthropic wants to create in AI. 3 month delay to get released. Randomly routed in biology, cyber and distillation. Weird sociopathic quiet harmfulness in ml topics. Goes away on June 22. Have to apply to get access to 'real' Mythos. What an exhausting tiresome dystopia.

kache avatar
⟳ nano retweeted
kache@yacineMTB·Jun 9

just so you guys are noticing this; they will pull the ladder from above you as soon as they can. their intentions are to disempower you as much as they reasonably can. the only reason they have given you anything at all is because openai has forced them to

Xin Eric Wang avatar
⟳ nano retweeted
Xin Eric Wang@xwang_lk·Jun 9

Motion for AI researchers: DO NOT evaluate and report results on Fable 5 models. Closeness hurts science, and gated access will further destroy it. Embrace open source models.

When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT. Anthropic estimated that this would affect approximately 0.03% of traffic.

Post attachment
Anthropic avatar
⟳ nano retweeted
Anthropic@AnthropicAI·May 7

New Anthropic research: Natural Language Autoencoders. Models like Claude talk in words but think in numbers. The numbers—called activations—encode Claude’s thoughts, but not in a language we can read. Here, we train Claude to translate its activations into human-readable text.

3:16
DeepSeek avatar
⟳ nano retweeted
DeepSeek@deepseek_ai·Apr 24

🚀 DeepSeek-V4 Preview is officially live & open-sourced! Welcome to the era of cost-effective 1M context length. 🔹 DeepSeek-V4-Pro: 1.6T total / 49B active params. Performance rivaling the world's top closed-source models. 🔹 DeepSeek-V4-Flash: 284B total / 13B active params. Your fast, efficient, and economical choice. Try it now at chat.deepseek.com via Expert Mode / Instant Mode. API is updated & available today! 📄 Tech Report: huggingface.co/deepseek-ai/D… 🤗 Open Weights: huggingface.co/collections/d… 1/n

Post attachment
METR avatar
⟳ nano retweeted
METR@METR_Evals·Apr 10

We ran GPT-5.4 (xhigh) on our tasks. Its time-horizon depends greatly on our treatment of reward hacks: the point estimate would be 5.7hrs (95% CI of 3hrs to 13.5hrs) under our standard methodology, but 13hrs (95% CI of 5hrs to 74hrs) if we allow reward hacks.

Post attachment
nano avatar
nano@nanulled·Mar 8

I seriously think that openai started purposely hurting ml research capabilities with this model, it's literally worse at taste than 5.2 high. I understand the competitive advantage of withholding capabilities but still they should just admit it and not waste anyone's time.

nano avatar
nano@nanulled·Mar 8

5.4 xhigh is worse than 5.3 codex at ml research, running experiments, patching gated features and debugging inference and evals. It's maybe better at moonshooting proposals just like 5.2 high but it does not have a robust experimentation hygiene. Same with 5.4 pro vs 5.2 pro.

Google DeepMind avatar
⟳ nano retweeted
Google DeepMind@GoogleDeepMind·Jan 29

Step inside Project Genie: our experimental research prototype that lets you create, edit, and explore virtual worlds. 🌎

2:03
Google DeepMind avatar
⟳ nano retweeted
Google DeepMind@GoogleDeepMind·5 Aug 2025

What if you could not only watch a generated video, but explore it too? 🌐 Genie 3 is our groundbreaking world model that creates interactive, playable environments from a single text prompt. From photorealistic landscapes to fantasy realms, the possibilities are endless. 🧵

2:23