Kristian Bonnici (@kristiandaaniel) header
Kristian Bonnici (@kristiandaaniel) profile picture

New here

Espoo, Suomi
Joined August 2017
Photo by Kristian Bonnici (@kristiandaaniel)Photo by Kristian Bonnici (@kristiandaaniel)Photo by Kristian Bonnici (@kristiandaaniel)Photo by Kristian Bonnici (@kristiandaaniel)Photo by Kristian Bonnici (@kristiandaaniel)
T3 Code was already the best ADE. now it's also the best looking one thanks to custom themes
This has been out for a bit now. Show me what you’ve built with it? 👀👀
2
4
67,928
Raw weights were supposed to be uninterpretable. that assumption is starting to crack
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math. It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training. We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵 Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math. rikiyat.github.io/blog/weight-sp… github.com/RikiyaT/lm_und… huggingface.co/Qwen/Qwen3.5-0… huggingface.co/RikiyaT/Qwen3.…
2
46
5h limit is GONE on Ultra (Max 20x)!! but lowkey feels like they're just prepping us for Fable 5.5's appetite 💀
We’re renaming our Max plans: Max 5x → Max Max 20x → Ultra Max is still $100/month, and Ultra is still $200. If you’re a subscriber, there’s nothing you need to do. We’re also removing the 5-hour limit for verified Ultra accounts, starting today with most existing subscribers. Newer accounts keep a temporary limit until they’re verified, which helps us prevent fraud and abuse. You'll get an email when it's removed. Weekly limits remain unchanged. View our plans here: claude.com/pricing
56
GPT-6.1 Terra is so underrated. Fastest computer use in Codex, half the price of Sol, and apparently the least likely to make stuff up
GPT-6.1 Terra is now OpenAI's least likely model to make things up
1
42
This! Every orchestration run I do ends with an HTML breakdown of the agents, tasks, and timeline in @t3dotcodes. Can’t imagine ever going back to anything less than this
I'll never get over how good HTML rendering looks in T3 code!
1
90
Built a lead-agent workflow on top of @theo's T3 Code: it coordinates Claude Code and Codex across my subscriptions. One approval to start, then separate threads and worktrees for each task, quota-aware pauses, and verified merges. Here it is in action. 🧵
Claude Managed Agents dynamic workflows are now available in public beta. It's a new type of multiagent orchestration, built for your most ambitious workloads. A lead agent writes a plan that runs across many agents in phases, combining the results at the end.
1
8
528
It checks git instead of trusting a task's report, merges, and launches what it unblocked. Inbox stays calm: the orchestrator pins itself to the top, settles each task once merged, and leaves only what needs me. With T3's Working section on, running agents stay folded away.
1
1
43
At the end: a report with a dependency graph and a timeline. Heavily customized to my workflow, so adapt it or use it as inspiration for your own skill: github.com/kristianbonnici/a…
42