(The funny thing is that anyone who actually is familiar with AI writing would know that the way I used em-dashes was very non-Claude-y)
LLMS
-
Latent Reasoning Tuning: LLMs Reasoning Without Full Step Output
By
–
What if LLMs could reason without writing out every single step? Researchers from Harbin Institute of Technology propose a new method called Latent Reasoning Tuning (LRT). Instead of generating long, costly text chains, LRT teaches the model to use compact internal
-
Speculative Decoding Optimization for Mixture of Experts Models
By
–
Get more from speculative decoding in MoE models
-
Local AI Models Cannot Match Leading Labs Models
By
–
Nah, never found a single local model that can replace the best labs’ models.
-
MoE-Based LLMs Enhance Speculative Decoding Effectiveness
By
–
New Technical Report from @EkagraRanjan
: Contrary to what you might expect, MoE-based LLMs make speculative decoding even more effective. Read more on our blog: -
Understanding AI Models: Working With Their Natural Strengths
By
–
The model has a grain. Most "bad output" is me trying to force it across that grain instead of along it.
-
How to Enable Summarized Thinking Mode in AI API Settings
By
–
OK, here's a resolution – I managed to get it to think using these settings: "thinking": { "type": "adaptive", "display": "summarized"
},
"output_config": { "effort": "max"
} Without "display": "summarized" I couldn't tell if it thought or not -
AI Model API No Longer Reports Reasoning Tokens Separately
By
–
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 -
Reasoning Tokens Not Reported Separately in API Response
By
–
Thanks, that did work! https://
gist.github.com/simonw/0f1a370
92fd0814bae75050eced48889#raw-json-response
… – the thing that confused me most is that it doesn't report "reasoning" tokens as a separate line item any more, see comment at the bottom of my Gist -
LLMs for Permission Control and Secret Blocking
By
–
We been testing llms to judge permissions, and proxies to block secrets. Interesting approach!