[HN Gopher] Show HN: PILF, The ultimate solution to catastrophic...
___________________________________________________________________
Show HN: PILF, The ultimate solution to catastrophic oblivion on AI
models
Author : NetRunnerSu
Score : 22 points
Date : 2025-06-27 11:10 UTC (11 hours ago)
(HTM) web link (github.com)
(TXT) w3m dump (github.com)
| Ifkaluva wrote:
| It's an interesting idea, I have two questions.
|
| - Surprise is detected by the norm of the gradients. So, doesn't
| this suggest that the model already has a way of adjusting to
| surprise?
|
| - Is there a danger of model instability when the gradients
| become larger and the learning rate is also increased?
| NetRunnerSu wrote:
| 1. an overly strong surprise is like PTSD in humans - it
| changes the model's previously learned experience forever, this
| is what we want to avoid
|
| 2. it's bound to happen, and our PILR-S is designed to keep the
| learning rate within the bell curve and decreasing as the
| surprise decreases (less new information, less learning).
| upghost wrote:
| This looks absolutely fantastic, please accept my meagre
| professional jealousy. I have long bemoaned manual hyperparam
| fiddling . I have on occasion dabbled with nonparametric
| ("genetic") methods of hyperparam tuning inspired by AutoML...
| but then you still have to manually tune the evolutionary
| hyperparams.
|
| Finding a way to derive this from the gradients is amazing.
| hackingonempty wrote:
| Parameters I'd Like to Fiddle
___________________________________________________________________
(page generated 2025-06-27 23:01 UTC)