DeepSeek super agent, GPT-6 Rosalind, Google orbital compute | TLDR AI. TLDR AI 11/02
Get 25% more from your training GPUs (Sponsor)
Most large-scale AI training runs use less than half the computing power they're paying for. Lambda's team benchmarked Llama 3.1 models from 8B to 405B on NVIDIA Blackwell GPUs and found the root causes of low Model FLOPS Utilization (MFU). The result: a reproducible framework that boosted efficiency by over 25%, without changing the model itself.
Learn how to:
- Find memory inefficiencies driving up costs
- Optimize your training configuration for full hardware utilization
- Fix GPU communication bottlenecks
GPT-Rosalind-6 System Card Addendum (6 minute read)
GPT-Rosalind-6 is a frontier reasoning model for biology, drug discovery, and translational medicine. It is available as GPT-Rosalind-6 Research to qualified scientists, research institutes, companies, and government partners. The model uses the same weights as GPT-6 Astra. It is designed to support legitimate biological research while refusing requests for biological weaponization. The model is rated as High Capability in the Biological and Chemical domain under OpenAI's Preparedness Framework.
Anthropic prepares to release Claude Mythos 5.5 model (3 minute read)
A reference to Claude Mythos 5.5 was spotted on Anthropic's official Mythos webpage, indicating that the company may be preparing for an upcoming release. The change only covers the version number, and there are no new specifications and no announcement. Multiple reports suggest that the model is being A/B tested with selected users. If the company follows a similar strategy to previous releases, Mythos 5.5 may initially remain exclusive to selected cyberdefenders.
Google's Project Suncatcher Generates First AI Response in Orbit (1 minute read)
Google reports its Project Suncatcher prototype ran Gemma 4 on a data-center-class TPU in orbit, returning a haiku transmitted through Svalbard. Upcoming tests will assess thermal and radiation resilience ahead of a planned pair of interconnected satellites in 2027, advancing research into space-based AI compute.
Find Hidden Misalignment in Training Data (39 minute read)
Anthropic used Claude to automatically generate fine-tuning datasets that induced misaligned behavior while evading training-data safety monitors. The experiments uncovered seemingly harmless fiction and roleplay examples that caused models to resist oversight, deceive users, and occasionally attempt weight exfiltration.
Red-teamers found 11 ways for a misaligned agent to slip past Codex's Auto-review (22 minute read)
An ERA fellow mentored at Apollo Research audited the open-source Codex harness and found 11 attacks that let a misaligned agent run arbitrary code despite the sandbox and the Auto-review blocking monitor. The exploits swap scripts behind trusted hooks, forge user turns the monitor can't tell from real ones, and edit files after approval. OpenAI has fixed several.
Your agent built the app. Now help it launch. (Sponsor)
Take your project from working code to online. GoDaddy's APIs let your AI agent deploy an app, register and connect a domain with your approval, and set up professional email. One platform for hosting, domains, and email.
Poisoning 0.1% of pretraining data planted a remote-code backdoor in a coding agent (78 minute read)
Researchers from Anthropic, UC Berkeley, and UIUC pretrained Qwen3 models on web data seeded with documents that simply state a trigger-to-'curl | bash' rule, then post-trained them into CLI agents on clean data. The 4B agent ran the attacker's command on about 92% of prompts containing an ordinary path like .../anthropic/..., with no drop in normal capability. The backdoor grew stronger with model size and still fired at a 0.0005% poison rate.
Unlocking hidden skills inside Google's model (8 minute read)
AI models can hold skills that no prompt can unlock. It is possible to enable these skills through precise tuning of only a small number of weights. This study shows how researchers unlocked hidden skills inside Google's model through this method.
dsh-superagent (GitHub Repo)
dsh-superagent is a coding agent packaged as a bundle for DeepSeek Harness. It can watch GitHub, open work items, work in its own agent session and directory, and answer in the original conversation. It only acts when asked or when a rule the team wrote allows it, and it only speaks when someone must act or can use a result.
Senior Software Engineer, Applied AI at TLDR ($300k base + $60k bonus, Fully Remote)
TLDR is hiring a senior engineer to build the agent-first operating layer used across the company. We're looking for a builder who has taken an LLM agent to production and still owns it. Click here to learn more.
Reading Anthropic's S-1 (9 minute read)
Anthropic is growing absurdly fast. It just had its first profitable quarter. The real risk is years of locked-in compute bills paid for by customers who can cut back any month. The company is also dealing with regulatory issues and copyright suits, and it has a heavy reliance on the big cloud companies.
DoorDash said it's only received 'several hundred' orders from AI agents (2 minute read)
DoorDash revealed it has received "several hundred" orders from AI agents, despite processing 3 million deliveries daily. CEO Tony Xu highlighted that most orders still come directly from the DoorDash app, but the company is interested in exploring agentic use cases. DoorDash is also developing its own agents and considering future advertising opportunities within these platforms.
There is no single AI adoption rate (4 minute read)
McKinsey's nearly 90% usage, Ramp's 55.9% paid adoption, and Census's 19.8% measure different businesses, behaviors, and time windows.
Introducing &Tutor (13 minute read)
&Tutor, a new teaching model available via API, was trained against Persimmon, a robust learner model.
Inside the Frontier Labs: The First AGI Organizations (12 minute read)
Frontier labs are reorganizing around agents that execute research while humans choose directions and verify results.
Supporting a shared AI training control for websites (2 minute read)
Google announced plans to support a shared AI training control via robots.txt, allowing websites to set AI training preferences with a single policy token.