3 17 7

Nikita Balagansky

elephantmipt

https://elephantmipt.github.io

AI & ML interests

None yet

Recent Activity

authored a paper 8 days ago

Next Embedding Prediction Makes World Models Stronger

upvoted a paper 9 days ago

Next Embedding Prediction Makes World Models Stronger

liked a Space 21 days ago

t-tech/manifolds

View all activity

Organizations

authored a paper 8 days ago

Next Embedding Prediction Makes World Models Stronger

Paper • 2603.02765 • Published 10 days ago • 19

upvoted a paper 9 days ago

Next Embedding Prediction Makes World Models Stronger

Paper • 2603.02765 • Published 10 days ago • 19

liked a Space 21 days ago

Chasing the Counting Manifold in Open LLMs

📚

Counting manifolds in open LLMs from behavior to SAEs.

updated a collection 22 days ago

Blogposts

Collection

1 item • Updated 22 days ago • 3

updated a Space 22 days ago

Chasing the Counting Manifold in Open LLMs

📚

Counting manifolds in open LLMs from behavior to SAEs.

published a Space 22 days ago

Chasing the Counting Manifold in Open LLMs

📚

Counting manifolds in open LLMs from behavior to SAEs.

upvoted a paper 22 days ago

Sanity Checks for Sparse Autoencoders: Do SAEs Beat Random Baselines?

Paper • 2602.14111 • Published 26 days ago • 55

liked a Space 29 days ago

Unlocking On-Policy Distillation for Any Model Family

📝

Visualize on-policy distillation for any model family

updated a Space about 1 month ago

Net2net

🚀

published a Space about 1 month ago

Net2net

🚀

liked a model 3 months ago

openai/circuit-sparsity

Text Generation • 0.4B • Updated Dec 12, 2025 • 687 • 203

upvoted a paper 3 months ago

T-pro 2.0: An Efficient Russian Hybrid-Reasoning Model and Playground

Paper • 2512.10430 • Published Dec 11, 2025 • 117

updated a dataset 4 months ago

elephantmipt/openwebtext_qwen3_tokenized

Viewer • Updated Nov 18, 2025 • 532k • 23

published a dataset 4 months ago

elephantmipt/openwebtext_qwen3_tokenized

Viewer • Updated Nov 18, 2025 • 532k • 23

upvoted a paper 4 months ago

Tongyi DeepResearch Technical Report

Paper • 2510.24701 • Published Oct 28, 2025 • 101

New activity in t-tech/flex-sae 5 months ago

Update tag

#1 opened 5 months ago by

danieldk

updated a model 5 months ago

t-tech/flex-sae

Updated Oct 27, 2025 • 4 • 12

liked a model 5 months ago

t-tech/flex-sae

Updated Oct 27, 2025 • 4 • 12

updated a collection 5 months ago

Kernels

Collection

1 item • Updated Dec 20, 2025

published a model 5 months ago

t-tech/flex-sae

Updated Oct 27, 2025 • 4 • 12

Nikita Balagansky

AI & ML interests

Recent Activity

Organizations

elephantmipt's activity

Chasing the Counting Manifold in Open LLMs

Chasing the Counting Manifold in Open LLMs

Chasing the Counting Manifold in Open LLMs

Unlocking On-Policy Distillation for Any Model Family

Net2net

Net2net

Update tag