We experimented with training Claude on examples of safe behavior in scenarios like our evaluation. This had only a small effect, despite being similar to our evaluation. We got further by rewriting the responses to portray admirable reasons for acting safely.
LLMS
-
Training Claude to Understand and Correct Misaligned AI Behavior
By
–
We found that training Claude on demonstrations of aligned behavior wasn’t enough. Our best interventions involved teaching Claude to deeply understand why misaligned behavior is wrong. Read more:
-
Anthropic Research Eliminates Harmful Behavior in Claude 4
By
–
New Anthropic research: Teaching Claude why. Last year we reported that, under certain experimental conditions, Claude 4 would blackmail users. Since then, we’ve completely eliminated this behavior. How?
-
Codex: A Transformative Tool for All Computer Work, Not Just Coding
By
–
codex is for everyone — a transformative tool for all work done with a computer, not just coding
-
Codex Goal Feature for Long-Running Tasks
By
–
/goal in Codex is wild!
— Vaibhav (VB) Srivastav (@reach_vb) 8 mai 2026
Give Codex the mission.
Tell it what “done” looks like.
Let it keep going until it hits the end state.
Perfect for big refactors, migrations, retry loops, long-running experiments, and all the “don’t stop until this works” tasks.
Enable it in… pic.twitter.com/6VZKGHt7Ch/goal in Codex is wild! Give Codex the mission.
Tell it what “done” looks like.
Let it keep going until it hits the end state. Perfect for big refactors, migrations, retry loops, long-running experiments, and all the “don’t stop until this works” tasks. Enable it in -
Routing tasks to cheapest model that meets quality bar
By
–
What changed in May 2026: > Before: you picked one model and committed.
> After: you route tasks to the cheapest model that meets the quality bar. DeepSeek V4-Pro scores within 7-8 points of Claude Opus 4.7 on SWE-bench. At 1/7th the cost during promo. The prompting skill -
DeepSeek as cheap second opinion alongside Claude
By
–
The key insight most people miss: DeepSeek replacing Claude entirely is the wrong move. DeepSeek as a $0.14 second opinion running alongside Claude is the right move. Same thinking system. Dramatically different API bill.
-
DeepSeek V4 integrated with Claude Code, reduced cost, parallel usage
By
–
DeepSeek V4 now speaks Claude Code natively. $0.14 per million tokens vs $5.00 for Claude Opus 4.7. Here's the exact setup that runs DeepSeek as a second opinion alongside your main Claude session, without replacing it: ———————————-
DEEPSEEK PARALLEL
— -
Representational Geometry Influences Uncertainty in Large Language Models
By
–
"Representational Curvature Modulates Behavioral Uncertainty in LLMs" Most interpretability work studies what features live inside LLMs. But this paper studies something deeper, which is the geometry of the whole evolving representation. The key result is that representational
-
LangChain adds API support for AI due diligence research agents
By
–
.
@p0
's API suite is now supported in LangChain, and they wrote a cookbook for a due diligence agent. We're part of the recipe:
• Deep Agents for orchestrating dedicated research track subagents
• @p0
’s Task API for research, findings, confidence scores, and follow-ups Full