SE-Bench: Benchmarking Self-Evolution with Knowledge Internalization Paper • 2602.04811 • Published 16 days ago • 2
iFSQ: Improving FSQ for Image Generation with 1 Line of Code Paper • 2601.17124 • Published 28 days ago • 32
Zero-Overhead Introspection for Adaptive Test-Time Compute Paper • 2512.01457 • Published Dec 1, 2025 • 2
Motion 3-to-4: 3D Motion Reconstruction for 4D Synthesis Paper • 2601.14253 • Published about 1 month ago • 10
UM-Text: A Unified Multimodal Model for Image Understanding Paper • 2601.08321 • Published Jan 13 • 9
ResTok: Learning Hierarchical Residuals in 1D Visual Tokenizers for Autoregressive Image Generation Paper • 2601.03955 • Published Jan 7 • 3
view post Post 6372 New family of 1B models just dropped!> LiquidAI/LFM2.5-1.2B-Base: 10T → 28T tokens> LiquidAI/LFM2.5-1.2B-Instruct: new large-scale multi-stage RL> LiquidAI/LFM2.5-1.2B-JP: our most polite model> LiquidAI/LFM2.5-VL-1.6B: multi-image multilingual> LiquidAI/LFM2.5-Audio-1.5B: 8x times faster, no quality lossSuper proud of this release 🤗 See translation 3 replies · 🚀 15 15 👀 1 1 + Reply
FlowBlending: Stage-Aware Multi-Model Sampling for Fast and High-Fidelity Video Generation Paper • 2512.24724 • Published Dec 31, 2025 • 7
Dream2Flow: Bridging Video Generation and Open-World Manipulation with 3D Object Flow Paper • 2512.24766 • Published Dec 31, 2025 • 9
What matters for Representation Alignment: Global Information or Spatial Structure? Paper • 2512.10794 • Published Dec 11, 2025 • 9
view post Post 855 Introducing StereoSpace -- our new end-to-end method for turning photos into stereo images without explicit geometry or depth maps. This makes it especially robust with thin structures and transparencies. Try the demo below:🌐 Project: prs-eth/stereospace_web📕 Paper: StereoSpace: Depth-Free Synthesis of Stereo Geometry via End-to-End Diffusion in a Canonical Space (2512.10959)🐙 Code: https://github.com/prs-eth/stereospace🤗 Demo: toshas/stereospace🤗 Weights: prs-eth/stereospace-v1-0By ETH Zürich ( @behretj , @Bingxin , @konradschindler ), University of Bologna ( @fabiotosi92 , @mpoggi ), HUAWEI Bayer Lab ( @toshas ). See translation 😎 3 3 + Reply
Error-Free Linear Attention is a Free Lunch: Exact Solution from Continuous-Time Dynamics Paper • 2512.12602 • Published Dec 14, 2025 • 44
StereoSpace: Depth-Free Synthesis of Stereo Geometry via End-to-End Diffusion in a Canonical Space Paper • 2512.10959 • Published Dec 11, 2025 • 12