Reading, Not Thinking: Bridging the Modality Gap When Text Becomes Pixels

Kaiser Sun, Xiaochuang Yuan, Hongjun Liu, Chen Zhao and 3 more

Atlanta Poster Session 5, Fri, Dec 11, 10:00 AM–1:00 PM, Hall C1 · Published 2026

Spin-Weighted Spherical Harmonics Enable Complete and Scalable E(3)-Equivariant Networks

Chenxing Liang, Yuchao Lin, Andrii Kryvenko, Wendi Yu and 4 more

Atlanta Poster Session 3, Thu, Dec 10, 10:00 AM–1:00 PM, Hall C1 · Published 2026

PerturbReason: A Knowledge-Grounded Benchmark and Framework for Cell-State–Conditioned Mechanistic Reasoning of Perturbation Effects

Dongkwan Kim, Yiming Gao, Yining Yang, Yang Shen

Atlanta Poster Session 6, Fri, Dec 11, 4:30 PM–7:30 PM, Hall C1 · Published 2026

AnaDiffusion: Anatomically Compositional Latent Diffusion for Controllable 3D Brain MRI Generation

Tracy Han, Lulin Liu, Bangya Liu, Yuanhao Cai and 7 more

Atlanta Poster Session 4, Thu, Dec 10, 4:30 PM–7:30 PM, Hall C1 · Published 2026

Tune-Up Open-Weight CLIP: Optimization Framework for Self-Supervised Fine-tuning of CLIP

Anant Mehta, Xiyuan Wei, Xingyu Chen, Tianbao Yang

Atlanta Poster Session 1, Wed, Dec 9, 10:00 AM–1:00 PM, Hall C1 · Published 2026

CLUE: Correlated Latent Uncertainty for Single-Pass Deep Uncertainty Estimation

Yucheng Wang, Wenyuan Zhao, Weifeng Zhang, Chao Tian and 1 more

Atlanta Poster Session 2, Wed, Dec 9, 4:30 PM–7:30 PM, Hall C1 · Published 2026

DRTriton: Large-Scale Synthetic Data Driven Reinforcement Learning for Triton Kernel Generation

DRTriton trains LLMs via synthetic data and reinforcement learning to generate optimized Triton kernels, achieving 92% speedup on KernelBench Level 2 versus 23% for GPT-5.2.

Siqi Guo, Ming Lin, Tianbao Yang

Atlanta Poster Session 2, Wed, Dec 9, 4:30 PM–7:30 PM, Hall C1 · Published 2026

Curvature Beyond Positivity: Greedy Guarantees for Arbitrary Submodular Functions

Curvature is extended to arbitrary submodular functions, yielding greedy multiplicative approximation guarantees that apply even to negative-valued objectives.

Yixin Chen, Alan Kuhnle

Atlanta Poster Session 1, Wed, Dec 9, 10:00 AM–1:00 PM, Hall C1 · Published 2026

DarkVGGT: Seeing Through Darkness Using Thermal Geometry without Daylight Tax

DarkVGGT uses physics-aware thermal modeling and geometry-shared routing to boost feed-forward 3D reconstruction in darkness without impairing daylight performance.

Minseong Kweon, Wenyuan Zhao, Nuo Chen, Lulin Liu and 5 more

Atlanta Poster Session 1, Wed, Dec 9, 10:00 AM–1:00 PM, Hall C1 · Published 2026

Not Another Text Benchmark: Putting the “Visual" Back in Visual Question Answering for Large Video Models

Three visual benchmarks for video understanding expose large video model weaknesses when reasoning through visual queries instead of text options.

Rwiddhi Chakraborty, Yinong O Wang, Cheng Zhang, Fan Bai and 5 more

Paris Poster Session 5, Fri, Dec 11, 11:30 AM–1:30 PM, Paris Poster Hall · Published 2026

Inference-Time Search Using Side Information for Diffusion-Based Image Reconstruction

A training-free inference-time search framework incorporates side information into diffusion-based inverse problem solvers to consistently improve reconstruction quality across diverse tasks.

Mahdi Farahbakhsh, Vishnu Teja Kunde, Dileep Kalathil, Krishna Narayanan and 1 more

Atlanta Poster Session 1, Wed, Dec 9, 10:00 AM–1:00 PM, Hall C1 · Published 2026

Attention Sinks and Outliers in Attention Residuals

OASIS stabilizes dual-normalized attention-residual architectures via null routing and token-to-depth null coupling, reducing activation outliers by 81.75% and improving low-bit quantized reasoning by 42.11%.

Haozheng Luo, Haoran Dai, Shaoyang Zhang, Xi Chen and 9 more

Atlanta Poster Session 6, Fri, Dec 11, 4:30 PM–7:30 PM, Hall C1 · Published 2026 · ▲ 1 on Hugging Face · Code ★ 3

DAGent: Evaluate-then-Grow Planning for Deep Research Agents

DAGent enables evaluate-then-grow DAG planning for deep-research agents, improving benchmarks by 2, 6 points over plan-then-patch baselines with lower token cost.

Hanwen Liu, Yuanfu Sun, Qiaoyu Tan

Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026 · ▲ 4 on Hugging Face · Code ★ 2

Reinforcement Learning for Diffusion LLMs with Entropy-Guided Step Selection and Stepwise Advantages

Formulating diffusion LLM generation as a finite-horizon MDP yields an exact stepwise policy gradient with entropy-guided step selection and one-step advantages, achieving state-of-the-art RL post-training results.

Vishnu Teja Kunde, Fatemeh Doudi, Mahdi Farahbakhsh, Dileep Kalathil and 2 more

Atlanta Poster Session 2, Wed, Dec 9, 4:30 PM–7:30 PM, Hall C1 · Published 2026