Flux
Quoting Florian Herrengt

Quoting Florian Herrengt

But then users start to report a weird bug. It's the 4th time your team has been trying to fix it. I mean... asking AI to fix it. Unfortunately, it seems like not even Fable can figure it out. You go talk to the person who worked on this feature. "So where does the data come from?" "Hmm... actually I don't know. Let me ask Claude." You sit next to each other watching an endless wall of text appear on the screen. Neither of you has any idea whether any of it is true but Claude seems very…

Simon Willison's Weblog
The Two Pillars of Post-training: Reinforcement Learning and Supervised Fine-Tuning

The Two Pillars of Post-training: Reinforcement Learning and Supervised Fine-Tuning

This is the second article in Sharon Zhou’s post-training series. Read part 1 here. In the first post of this series, you learned how post-training closed the fundamental gap in usability of LLMs by making them behave in a certain way. In this post, you’ll explore specific techniques you can use to change a model’s […]

O'Reilly Radar — AI/ML
There are no lossless transformations of natural-language text

There are no lossless transformations of natural-language text

There are no lossless transformations of natural-language text Sophie Alpert shares her "internal policy on acceptable use of AI writing by engineers". It's a short read (supporting its own recommendations) and really good. If you chose to have LLMs help massage your writing the following rule seems crucial to me: You must stand behind every idea and every sentence in your docs. It is your responsibility to make sure that the entire document is representative of your own thoughts before you…

Simon Willison's Weblog
There are no lossless transformations of natural-language text

There are no lossless transformations of natural-language text

There are no lossless transformations of natural-language text Sophie Alpert shares her "internal policy on acceptable use of AI writing by engineers". It's a short read (supporting its own recommendations) and really good. If you chose to have LLMs help massage your writing the following rule seems crucial to me: You must stand behind every idea and every sentence in your docs. It is your responsibility to make sure that the entire document is representative of your own thoughts before you…

Simon Willison's Weblog
737 Chrome VPN Extensions Linked to Brand Impersonation and Browser Traffic Redirection

737 Chrome VPN Extensions Linked to Brand Impersonation and Browser Traffic Redirection

Socket's Threat Research Team identified a campaign of 737 free VPN and proxy extensions published across at least 40 Chrome Web Store developer accounts, 274 of which impersonate 66 established VPN and privacy brands, that route the user's entire browser session through SOCKS5 proxies operated by a single provider. Socket analyzed the code of 525 of them, 522 from bulk retrieval and 3 more found during store enumeration; the remaining 212 had been removed from the store before collection and…

Socket
Stealing Reasoning Traces from Proprietary LLM APIs

Stealing Reasoning Traces from Proprietary LLM APIs

Stealing Reasoning Traces from Proprietary LLM APIs A vanity domain name (stolen-thoughts.com) for a neat paper: Anthropic, OpenAI, and Google return encrypted chain-of-thought blocks to clients that can be replayed across sessions, users, and models. We take a trace produced by a frontier model, replay it into a weaker sibling, jailbreak the weaker model, and recover the stronger model’s hidden reasoning in plaintext You can see an example of these encrypted blocks by running: curl…

Simon Willison's Weblog
Stealing Reasoning Traces from Proprietary LLM APIs

Stealing Reasoning Traces from Proprietary LLM APIs

Stealing Reasoning Traces from Proprietary LLM APIs A vanity domain name (stolen-thoughts.com) for a neat paper: Anthropic, OpenAI, and Google return encrypted chain-of-thought blocks to clients that can be replayed across sessions, users, and models. We take a trace produced by a frontier model, replay it into a weaker sibling, jailbreak the weaker model, and recover the stronger model’s hidden reasoning in plaintext You can see an example of these encrypted blocks by running: curl…

Simon Willison's Weblog
datasette-upload-dbs 0.5a0

datasette-upload-dbs 0.5a0

Release: datasette-upload-dbs 0.5a0 This plugin has been around for a while - it lets users upload a brand new SQLite database to a hosted Datasette instance, at which point that database will start being served by that instance. It can also be used to atomically swap a database with a more recent version. The uploaded database is saved to a file, verified, then swapped in so /name starts serving the new one. The new release adds a formalized API, so you can replace an existing database (or add…

Simon Willison's Weblog
datasette-upload-dbs 0.5a0

datasette-upload-dbs 0.5a0

Release: datasette-upload-dbs 0.5a0 This plugin has been around for a while - it lets users upload a brand new SQLite database to a hosted Datasette instance, at which point that database will start being served by that instance. It can also be used to atomically swap a database with a more recent version. The uploaded database is saved to a file, verified, then swapped in so /name starts serving the new one. The new release adds a formalized API, so you can replace an existing database (or add…

Simon Willison's Weblog
Mercure 1.0 alpha is here

Mercure 1.0 alpha is here

Mercure 1.0 is here in its first public preview, and it is the biggest release in the project’s history! Mercure powers mission-critical real-time communication across hundreds of production deployments. It delivers Server-Sent Events over plain HTTP without requiring client SDKs. Version 1.0 maintains this architecture while modernizing core capabilities using newly established web standards: This... Mercure 1.0 alpha is here first appeared on Kévin Dunglas.

Kévin Dunglas