euclaise

euclaise's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

upvoted a paper 1 day ago

From Next-Token to Next-Block: A Principled Adaptation Path for Diffusion LLMs

Paper • 2512.06776 • Published 5 days ago • 20

upvoted 2 papers 2 days ago

ARC-AGI Without Pretraining

Paper • 2512.06104 • Published 6 days ago • 1

Beyond Real: Imaginary Extension of Rotary Position Embeddings for Long-Context LLMs

Paper • 2512.07525 • Published 4 days ago • 52

upvoted an article 3 days ago

view article

Article

The 1 Billion Token Challenge: Finding the Perfect Pre-training Mix

• 

Nov 3

•

51

upvoted 16 papers 3 days ago

The Path Not Taken: RLVR Provably Learns Off the Principals

Paper • 2511.08567 • Published about 1 month ago • 32

To Think or Not to Think: The Hidden Cost of Meta-Training with Excessive CoT Examples

Paper • 2512.05318 • Published 7 days ago • 2

A Hitchhiker's Guide to Scaling Law Estimation

Paper • 2410.11840 • Published Oct 15, 2024 • 1

Reasoning Under 1 Billion: Memory-Augmented Reinforcement Learning for Large Language Models

Paper • 2504.02273 • Published Apr 3 • 7

Structured Linear CDEs: Maximally Expressive and Parallel-in-Time Sequence Models

Paper • 2505.17761 • Published May 23 • 1

Fixed-Point RNNs: Interpolating from Diagonal to Dense

Paper • 2503.10799 • Published Mar 13 • 1

Weight-Space Linear Recurrent Neural Networks

Paper • 2506.01153 • Published Jun 1 • 1

Scaling Linear Attention with Sparse State Expansion

Paper • 2507.16577 • Published Jul 22 • 1

The Automated LLM Speedrunning Benchmark: Reproducing NanoGPT Improvements

Paper • 2506.22419 • Published Jun 27 • 15

A Minimalist Optimizer Design for LLM Pretraining

Paper • 2506.16659 • Published Jun 20 • 1

Tiled Flash Linear Attention: More Efficient Linear RNN and xLSTM Kernels

Paper • 2503.14376 • Published Mar 18 • 1

Masked Gated Linear Unit

Paper • 2506.23225 • Published Jun 29 • 1

WPO: Enhancing RLHF with Weighted Preference Optimization

Paper • 2406.11827 • Published Jun 17, 2024 • 16

Selective Rotary Position Embedding

Paper • 2511.17388 • Published 20 days ago • 1

Deep Improvement Supervision

Paper • 2511.16886 • Published 21 days ago • 1

ReGLA: Refining Gated Linear Attention

Paper • 2502.01578 • Published Feb 3 • 1