TWN attaches separate reasoning and embedding LoRA adapters to a frozen backbone with gradient detachment and a self-supervised gate that decides per input whether to generate CoT, achieving SOTA on MMEB-V2 with 3-5% added parameters and up to 50% fewer reasoning tokens.
Learning transferable visual models from natural language supervision
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
method 1
citation-polarity summary
years
2026 3verdicts
UNVERDICTED 3roles
method 1polarities
use method 1representative citing papers
Diffusion models can extract reusable density-mode concepts from their time-indexed scores to enable compositional generation at test time on held-out benchmarks from ColorMNIST and CelebA.
SCOUP decouples 2D sparse code learning from 3D Gaussian optimization to deliver up to 400x training speedup and 3x better memory efficiency while matching accuracy on open-vocabulary 3D queries.
citing papers explorer
-
Think When Needed: Adaptive Reasoning-Driven Multimodal Embeddings with a Dual-LoRA Architecture
TWN attaches separate reasoning and embedding LoRA adapters to a frozen backbone with gradient detachment and a self-supervised gate that decides per input whether to generate CoT, achieving SOTA on MMEB-V2 with 3-5% added parameters and up to 50% fewer reasoning tokens.
-
Test-Time Compositional Generalization in Diffusion Models via Concept Discovery
Diffusion models can extract reusable density-mode concepts from their time-indexed scores to enable compositional generation at test time on held-out benchmarks from ColorMNIST and CelebA.
-
Sparse Code Uplifting for Efficient 3D Language Gaussian Splatting
SCOUP decouples 2D sparse code learning from 3D Gaussian optimization to deliver up to 400x training speedup and 3x better memory efficiency while matching accuracy on open-vocabulary 3D queries.