Maxim Bobrin @maxsbob21

Novelty Search & Neural Evolution Q: "What makes someone truly creative? What is the main driving force?" BSc & MSc in Mathematics Joined April 2025

Tweets

24
Followers

51
Following

135
Likes

661

Maxim Bobrin @maxsbob21

a day ago

@zhaofeng_wu bros reporting <0.1 loss reduction lol

0 0 0 450 0

View Details

Maxim Bobrin @maxsbob21

3 days ago

@ChenTessler yes, but it would be interesting also to infer behavior from text description, e.g run as fast as you can and make a backflip at the end

0 0 0 5 0

View Details

ICML 2026: Latent Reasoning in TRMs is Secretly a Policy Improvement Operator Why does recursive reasoning, especially latent reasoning, actually work? The theory is still young, and even mechanistic explanations are limited. We close part of this gap by showing that latent reasoning is secretly doing policy improvement. Each recursion pushes the model steadily toward the target. Based on this view, we propose an algorithm that boosts learning and inference efficiency by up to 18x.

9 48 376 24K 397

View Details

Maxim Bobrin @maxsbob21

5 days ago

@antirez bro got generational aura debt lol

0 0 0 49 0

View Details

Maxim Bobrin @maxsbob21

2 weeks ago

Lots of other interesting insights were made during this work. Thanks @machinestein, see everyone interested in Seoul!

Arip @machinestein

2 weeks ago

ICML 2026: Zero-Shot Off-Policy Learning Distribution ratios play a central role in off-policy reinforcement learning. We show that using only one behavioral foundational model, you can obtain its ratio to other policies for free, without any importance sampling or min-max

3 40 327 21K 267

1 0 4 194 0

View Details

Maxim Bobrin @maxsbob21

2 weeks ago

@zhaisf @geoffreyhinton Even without reading slides, the most probable explanation is that mnist contains all required features for predicting 3 based on other numbers (some of which look similar to 3). It would be more interesting to take most non representative digit and check OOD

2 0 9 953 0

View Details

Maxim Bobrin @maxsbob21

4 weeks ago

@its_vayishu Am i correct that you store smth like a latent buffer that acts as a memory and based on this dynamic buffer predictive model outputs next states? And you show that there is some structure arising from memory?

1 0 1 45 0

View Details

Maxim Bobrin @maxsbob21

a month ago

@willccbb How is this different from Unsipervised Environment design? In RL there are already lots of papers and i assume some folks already managed to apply ideas from UED. Seems like this paper is just the same findings albeit from other perspective

2 0 0 189 0

View Details

Maxim Bobrin @maxsbob21

3 months ago

@Amank1412 Most ppl would say that they are optimizing their work, but in reality will spend more free time for soom scrolling

0 0 0 5 0

View Details

Maxim Bobrin @maxsbob21

3 months ago

@heyrimsha can we pla stop already? There are a lot of services for those kind of things… Research community needs to heal asap

2 0 0 321 0

View Details

Maxim Bobrin @maxsbob21

3 months ago

@daniel_mac8 they tested this on public games. Private is much harder

0 0 1 34 0

View Details

Maxim Bobrin @maxsbob21

3 months ago

@agenticasdk gg

0 0 0 86 0

View Details

Maxim Bobrin @maxsbob21

3 months ago

@KeyTryer You can check several replays and see that gemini is the only model which tries to understand what to do through reasoning traces, while other models just return action

0 0 0 279 0

View Details

Maxim Bobrin @maxsbob21

3 months ago

@loveofdoing How you came up with this approach? somekind of blogpost would be beneficial

0 0 0 581 0

View Details

Maxim Bobrin @maxsbob21

3 months ago

@ChenTessler Seems like there is no method that can do this even in simulation without spending hours on RL only for this single level

0 0 0 10 0

View Details

Maxim Bobrin @maxsbob21

3 months ago

@pengzhangzhi1 @Anru_Zhang @AlexanderTong7 Where to find recordings?

1 0 0 86 0

View Details

Maxim Bobrin @maxsbob21

3 months ago

@itsolelehmann what is 56%? How it is measured? Agent can basically just find some adverserial solution, get improvement on metric that he chose by himself and you will never notice. dafuq?