Skip to content
HN On Hacker News ↗

Can Bölük (@_can1357) on X

▲ 56 points 6 comments by himata4113 2w ago HN discussion ↗

Pangram verdict · v3.3

We believe that this entire text is human-written.

0 %

AI likelihood · overall

Human
100% human-written 0% AI-generated
SEGMENTS · HUMAN 1 of 1
SEGMENTS · AI 0 of 1
WORD COUNT 156
PEAK AI % 0% · §1
Analyzed
Aug 11
backend: pangram/v3.3
Segments scanned
1 windows
avg 156 words each
Distribution
100 / 0%
human / AI fraction
Verdict
Human
Pangram v3.3

Article text · 156 words · 1 segments analyzed

Human AI-generated
§1 Human · 0%

Can Bölük on X: "guys you do know you can just disable thinking, and instead give it a "deep_think" tool, and it will call it with internal CoT reasoning format right? gl fixing that https://t.co/eWnPGbwxXs"guys you do know you can just disable thinking, and instead give it a "deep_think" tool, and it will call it with internal CoT reasoning format right? gl fixing thatHere's a PoC pasta.can.ac/omegiligox.py Note how I said nothing about the format of the argument and we get gpt55+'s grug-talk format. Beyond this, which is kinda funny.kinda missed having proper thinking streams ngl. switching thinking levels work fine too, it has always been a number in the system prompt!will be up in next release, hf!fable btw, anthropic usually has great regex game, smhand I didn't even get a freaking domain for it! crazy!!!Now you've done it. Soon we will only be able to use these models in cloud-hosted harnesses provided by each lab.