It's a custom model that's on par with Maia 3 in terms of policy but estimates a lot more than that (e.g. difficulty).
Does this give sensible analysis if you don't record your thinking out loud? I've been using Elogram (https://elogram.gg/, made by a friend), which gives interpretability by showing how difficulty of a position changes…
As far as I understand: 1) it's very fast (they claim 40-200x faster than frontier models [1], would roughly line up with it doing diffusion) 2) each answer carries a calibrated probability (ie. frequency of outcome is…
Another hypothesis is that OpenAI's internal web sandbox (WebCache, also used during HF attack [1]) was too narrow/restrictive for the agents' purposes. During collusion-wiki, it was hypothesized that the sandbox would…
You need to separate memory capacity and bandwidth. Looping decreases memory capacity/FLOP but not bytes loaded/FLOP, since weights need to be loaded again for the 2nd pass. Plus (depending on the method used) capacity…
What LLM-specific hardware improvements should one expect? Seems to me that LLM inference is simple architecturally (matmul et al) so most scaling in hardware should come from general improvements (memory BW, packaging,…
Consider that a lot of the resistance to just existing is external expectations. I feel like I should always be doing something because I'm pressured to, or that small things bother me because I feel I need to perform.…
It's a custom model that's on par with Maia 3 in terms of policy but estimates a lot more than that (e.g. difficulty).
Does this give sensible analysis if you don't record your thinking out loud? I've been using Elogram (https://elogram.gg/, made by a friend), which gives interpretability by showing how difficulty of a position changes…
As far as I understand: 1) it's very fast (they claim 40-200x faster than frontier models [1], would roughly line up with it doing diffusion) 2) each answer carries a calibrated probability (ie. frequency of outcome is…
Another hypothesis is that OpenAI's internal web sandbox (WebCache, also used during HF attack [1]) was too narrow/restrictive for the agents' purposes. During collusion-wiki, it was hypothesized that the sandbox would…
You need to separate memory capacity and bandwidth. Looping decreases memory capacity/FLOP but not bytes loaded/FLOP, since weights need to be loaded again for the 2nd pass. Plus (depending on the method used) capacity…
What LLM-specific hardware improvements should one expect? Seems to me that LLM inference is simple architecturally (matmul et al) so most scaling in hardware should come from general improvements (memory BW, packaging,…
Consider that a lot of the resistance to just existing is external expectations. I feel like I should always be doing something because I'm pressured to, or that small things bother me because I feel I need to perform.…