https://
gist.github.com/simonw/0f1a370
92fd0814bae75050eced48889
… seemed to work – the problem is it no longer reports "reasoning" tokens as a separate line item from output tokens so I couldn't tell if reasoning had happened or not until I turned on the reasoning summary
RESEARCH
-
AI Model API No Longer Reports Reasoning Tokens Separately
By
–
-
Image Model Achieves Unprecedented Detail Quality
By
–
insane detail that I did not think possible in an image model in 10 years. congrats!!
-
LLMs Autonomously Develop Hash-like Random Number Extraction Algorithms
By
–
One of the most intriguing findings from the paper's analysis is that LLMs can autonomously devise random number extraction algorithms akin to hash functions (such as Sum-Mod or rolling hashes) within the context. The longer the inference model "thinks," the greater the
-
AI Models Now Good Enough After Last Year’s Limitations
By
–
Peekaboo was that but the models were just not good enough last year. Now they are.
-
Knowledge Cutoff Limitations in AI Models Explained
By
–
haha ouch, that is not malicious tho, just knowledge cutoff. Can’t blame them for that.
-
AI Correctly Identifies Redacted Federal Agency Communication From 15 Years Ago
By
–
Correctly bingoes a 3 paragraph communication to a federal agency written 15 years ago where all biographical hints were replaced with [REDACTED FOR PURPOSE OF EVAL].
-
Stanford researchers identify AI chatbot delusional spiral risks
By
–
AI chatbots can trap users in "delusional spirals," where chatbots affirm and amplify users' grandiose, paranoid, or imaginary beliefs without pushback. Stanford researchers identified key hallmarks and offer recommendations to address this problem:
-
Model Weights Alone Reveal Identity Without System Prompts
By
–
To head off the obvious question: This was conducted via the API and so it should not have hidden memory/system prompts/etc which leak the identity of yours truly to the model. The model weights, alone, got it to enough information.
-
Claude Opus Identifies Writer by Stylistic and Biographical Tells
By
–
Opus' rationale for the one it got right begins with the accurate and somewhat disconcerting "RATIONALE: Several strong stylistic and biographical tells point to Patrick McKenzie."
-
Claude Opus 4.7 Reproduces Writing Samples in Informal Test
By
–
FYI, I casually tried to reproduce this on three writing samples from many years ago, which I do not believe to be on the public Internet and which were varying subjective difficulty levels. Opus 4.7 went 1 for 3 with ~no effort in the prompt (~two sentences long).