Nathan Lambert
speaker
1,013 appearances
1 recordings
1 series
first heard Nov 2024
last heard Nov 2024
Nathan Lambert’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsNo recordings in the last 12 months.Older appearances are listed below; set an alert to hear about the next one.
Appearances
It bridges the gap between VLM and planner in a very nice way.
Where the VLM could be asked a question of like,
How do I or like what tool do I need to do X?
And then
It can just point at it rather than saying the answer.
Which like and then the planner knows to get things that are pointed at or something.
And that is interesting and some of my robotics friends are excited about it.
Yeah, it it's I mean it's very related to this kind of PPO DPO debate where a lot of the proponents of PPO is that you're
scoring and updating the model based on things that is generating itself rather than completions you got from elsewhere.
And that's really the thing.
It's just like it seems like it's a bit better learning signal if within these batches and these pairwise comparisons the
the tokens you are looking at are closer to the log problems of the current model.
Which I think
makes a lot of sense because what you do at like say DPO is you compute the log problems of the token in a completion and that updates your weight.
So it's like okay like
It makes sense if the model the log props are not in like a weird space it hasn't seen before.
And
It's nice to just see that it backed up by experiments.
Future.
It'll end up being fine.
Showing 941–960 of 1,013 · page 48 of 51
← Previous
Next →