Submissions by jasondavies

Seiler's Interpolation for Evaluating Polynomial Curves

Dragonfly: A large vision-language model with multi-resolution zoom

Knockout: A simple way to handle missing inputs

SaySelf: Teaching LLMs to Express Confidence with Self-Reflective Rationales

Intelligent Go-Explore: Standing on the Shoulders of Giant Foundation Models

Transformers are SSMs (Mamba-2)

Compressed-Language Models for Understanding Compressed File Formats: JPEG

RB-Modulation: Training-Free Personalization of Diffusion Models

Thermox: The First Thermodynamic Computing Simulator

Zipper: A Multi-Tower Decoder Architecture for Fusing Modalities

Claude can now use tools

Era3D: High-Resolution Multiview Diffusion Using Efficient Row-Wise Attention

Sonic: A Low-Latency Voice Model for Lifelike Speech

3DitScene: Editing Any Scene via Language-Guided Disentangled Gaussian Splatting

Geometry-Informed Neural Networks

Specular Polynomials

Grokked Transformers Are Implicit Reasoners

Kolmogorov-Arnold Networks (KANs) for Time Series Analysis

Look Once to Hear: Target Speech Hearing with Noisy Examples

MoRA: High-Rank Updating for Parameter-Efficient Fine-Tuning

StopThePop: Sorted Gaussian Splatting for View-Consistent Real-Time Rendering

Glyph-ByT5: A Customized Text Encoder for Accurate Visual Text Rendering

Generative Camera Dolly: Extreme Monocular Dynamic Novel View Synthesis

Thermodynamic Natural Gradient Descent

Octo: An Open-Source Generalist Robot Policy

Haptic Source-effector: Full-body Haptics via Non-invasive Brain Stimulation [pdf]

UK Semiconductor Institute Announced

Advanced AI evaluations at AISI: May update

CAT3D: Create Anything in 3D with Multi-View Diffusion Models