GraphForge: Training Working Agents with Graph-Anchored Workspace Synthesis

GraphForge synthesizes workspace tasks and verifiers over real file evidence graphs to train working agents, and fine-tuning Qwen3.6-27B improves GDPVal, Workspace-Bench-Lite, and SpreadsheetBench II results.

Qisheng Su, Hanchen Wang, 朱冠儒, Huicheng Jiang and 8 more

Published Sep 30, 2026 · 0 citations · ▲ 146 on Hugging Face

MinkowskiPE: Minkowski Positional Encoding for Spatiotemporal Perception

MinkowskiPE applies Lorentz transformations via joint spacetime positional encoding to make attention depend only on relative displacement, improving molecular dynamics and video prediction with far fewer parameters.

Yuhao Li, Louie Hong Yao, Tianyi Shi, Hanqun Cao and 3 more

Published Sep 27, 2026 · 0 citations · ▲ 14 on Hugging Face

Rethinking Token Reweighting for SFT: Suppress, Reverse, and Extrapolate Learned Features

Existing token-reweighting methods cannot reverse harmful SFT features; SCALE uses frozen SFT deltas with entropy-guided gates to suppress, reverse, or extrapolate them, improving math and code results.

Cunchun Li, Haonan He, Yifan Gao, Minglei Li and 3 more

Published Sep 27, 2026 · 0 citations · ▲ 11 on Hugging Face

OpenART: Scaling Agent Red Teaming via Open-Ended Environment Evolution

OpenART scales agent red teaming via open-ended environment evolution across 10,000 stateful scenarios, with EMHA achieving 85% attack success that grows with complexity.

Yunhao Chen, Xin Wang, Yixu Wang, Yi Liu and 5 more

Published Aug 1, 2026 · 0 citations · ▲ 266 on Hugging Face · Code ★ 231

CircuitSeer: Mining High-Quality Data by Probing Mathematical Reasoning Circuits in LLMs

Shaobo Wang, Yongliang Miao, YuanCheng Liu, Qianli Ma and 3 more

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

StreamPI: Streaming Multimodal Temporal Modeling for Vision-Language-Action Models

Zhe Liu, Jinghua Hou, Yuxiang Lu, Zhenya YANG and 6 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Do More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness

Songyuan Sui, Zhen Tan, Mohan Zhang, Rana M Khan and 2 more

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

MedVTok: A General-Purpose Medical Visual Tokenizer

Chenglong Ma, Yuanfeng Ji, Junzhi Ning, Jiyao Liu and 17 more

Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Object Hallucination Mitigation in Large Vision-Language Models via Self-Vision Dual Masking and Uncertainty-Triggered Assembly

Ziji Sheng, Guiyao Tie, Weidong Wang, Jiawen Shi and 7 more

Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

InfCoiL: Coordinated Planner-Controller Learning for Closed-Loop Physics-Based Human-Object Interaction

Yude Zou, Yifei Yao, Yiwei Hao, Hanqing Wang and 1 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

MMTA: Benchmarking Multimodal Temporal Analysis with Time Series, Text, and Vision

Ziyang Zhang, Shenyi Li, Yilin wang, Ziyun Cui and 3 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

From Perception to Punchline: Empowering VLM with the Art of In-the-wild Memes

Xueyan Li, Yingyi Xue, Mengjie Jiang, Qingzi Zhu and 1 more

Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

ABHBench: Evaluating Moral Decision-Making of Foundation Models from an Agentic Perspective

Chaoran Li, Yukun Li, Zeyuan Zhao, Shao Zhang and 2 more

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Does the Question Really Matter? Training-Free Data Selection for Vision-Language SFT

Peng Sun, Yi Yang, Huawen Shen, Yi Ban and 3 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Are Text-to-Image Models Inductivist Turkeys? A Counterfactual Benchmark for Causal Reasoning

jiayi lei, Yuandong Pu, Xingyu Han, Rongpeng Zhu and 7 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Penalize to Verify: A Multi-Domain Benchmark for Trustworthy LLM Evaluation

Jia-chen Zhang, Zheng Zhou, Zhuo Wang, Qi Jia and 6 more

Paris Poster Session 1, Wed, Dec 9, 12:30 PM–2:30 PM, Paris Poster Hall · Published 2026

$\text{A}^3$-VLA: Automatic Perception-Guided Attention Alignment for Robot Manipulation

Shengzhe Zhang, Qi Zhang, Dazhong Shen, Chao Wang and 2 more

Paris Poster Session 3, Thu, Dec 10, 12:30 PM–2:30 PM, Paris Poster Hall · Published 2026

BEAKER: An Expert-Curated Benchmark for Embodied Brains in Self-Driving Chemical Laboratories

Fei Lin, Tengchao Zhang, Ziyang Gong, Xiaotong Yu and 14 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Heterogeneity-aware Distillation for Federated Continual Learning

Gaozhuo Liu, Yichen Li, Xiuying Wang, Yulong Li and 3 more

Paris Poster Session 1, Wed, Dec 9, 12:30 PM–2:30 PM, Paris Poster Hall · Published 2026

Context Value Informed In-Context Reinforcement Learning

Wenhao Zhang, Shao Zhang, Xihuai Wang, Yang Li and 1 more

Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

IFDECORATOR: Wrapping Instruction Following Reinforcement Learning with Verifiable Rewards

Xu Guo, Tianyi Liang, Jian Tong, Xiaogui Yang and 5 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

PARE: Pruning and Adaptive Routing for Efficient Video Generation

PARE combines structure-aware width pruning and timestep-conditioned adaptive depth routing to cut video diffusion compute while preserving generation quality.

Yutong Wang, Yunke Wang, Tianfan Xue, Yu Qiao and 3 more

Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

PhotoFlow: Agentic 3D Virtual Photography Missions

PhotoFlow uses a Director-Reviewer-Reflector agent for closed-loop camera search to generate language-conditioned virtual photographs in arbitrary 3D scenes, outperforming baselines on quality, alignment, and success rate.

Jiarui Guo, Haojia Wei, Yiming Zhang, Yifei Liu and 4 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026 · ▲ 24 on Hugging Face · Code ★ 43

Experience Makes Skillful: Enabling Generalizable Medical Agent Reasoning via Self-Evolving Skill Memory

SkeMex improves medical agents via self-evolving skill memory that distills reusable procedural knowledge, governs retention by utility, and outperforms memory-based agents across clinical tasks.

Haoran Sun, Wenjie Li, Yujie Zhang, Zekai Lin and 7 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026 · ▲ 4 on Hugging Face

FinReasoning: A Hierarchical Benchmark for Reliable Financial Research Reporting

FinReasoning hierarchically benchmarks LLMs on financial research via semantic consistency, data alignment, and insight, revealing closed-source models suit core reasoning, open-source models lack consistency, and financial models lack auditing skills.

Yiyun Zhu, Yidong Jiang, Ziwen Xu, Yinsheng Yao and 3 more

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

ProCLIP: Progressive Vision-Language Alignment via LLM-based Embedder

ProCLIP progressively aligns CLIP image encoders with LLM-based embedders via curriculum distillation and contrastive tuning to support long multilingual texts without disrupting pretrained vision-language alignment.

Xiaoxing Hu, Kaicheng Yang, Ziqi Ye, Ziyang Gong and 6 more

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026 · ▲ 11 on Hugging Face · Code ★ 27

Meow-Omni 1: A Multimodal Large Language Model for Feline Ethology

Meow-Omni 1 integrates video, audio, physiological time-series, and text to reach 71.16% feline intent recognition, outperforming baseline multimodal models.

Jucheng Hu, Zhangquan Chen, Yulin Chen, Chengjie Hong and 8 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026 · ▲ 3 on Hugging Face

Defense-as-Skill: Evolving Runtime Guard Skill for Skill-Augmented Agents

Defense-as-Skill implements runtime guard SkillSonar as an editable skill that checks actions against task boundaries, reducing attack success rates substantially across agents via evolved guard-skill optimization.

Xiaofang Yang, Ziqi Miao, Dianbo Sui, Jing Shao and 1 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

SciHazard: A Benchmark for Measuring Scientific Safety Risks with Decomposed Harm Scoring

SciHazard benchmarks LLM scientific safety risks via decomposed harm scoring across 3,000 real-world grounded queries, finding deep research agents 32.3% more harmful than standard models.

Chunxiao Li, Yuan Xiong, Lijun Li, Tianyi Du and 3 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

AgenticOCR: Parsing Only What You Need for Efficient Retrieval-Augmented Generation

AgenticOCR transforms OCR into query-driven, on-demand extraction to selectively parse document regions, improving visual RAG efficiency and accuracy over page-level chunking.

Zhengren Wang, Dongsheng Ma, Huaping Zhong, Jiayu Li and 5 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

AOT-POT: Adaptive Operator Transformation for Large-Scale PDE Pre-training

AOT-POT adaptively transforms diverse PDE solution operators into simpler aligned forms via multi-stream aggregation and Sinkhorn mixing, achieving state-of-the-art pre-training accuracy with minimal parameters.

Qitan Lv, Hong Wang, Hao Zhongkai, Wen Wu and 4 more

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

BrainVista: Modeling Naturalistic Brain Dynamics as Multimodal Next-Token Prediction

BrainVista models brain dynamics via multimodal next-token prediction with network tokenizers and stimulus masking, achieving state-of-the-art fMRI encoding and up to 36% better long-horizon rollout correlations.

Xuanhua Yin, Ted Zhao, Lina Yao, Weidong Cai

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Leveraging Error Diversity in Group Rollouts for Reinforcement Learning

RLVR discards group error diversity, but shaping penalties by intra-group error diversity improves training; EDAS boosts DAPO by 6.29 points on math benchmarks.

Wenpu Liu, Yuqi Xu, Weichu Xie, Yongfu Zhu and 7 more

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Visual-ERM: Reward Modeling for Visual Equivalence

Visual-ERM is a multimodal generative reward model that evaluates vision-to-code outputs in rendered visual space, improving Qwen3-VL-8B-Instruct by up to 8.4 points and outperforming larger models on fine-grained visual discrepancy benchmarks.

Ziyu Liu, Shengyuan Ding, Xinyu Fang, Xuanlang Dai and 5 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026 · ▲ 21 on Hugging Face · Code ★ 69

PLANING: A Loosely Coupled Triangle-Gaussian Framework for Streaming 3D Reconstruction

PLANING decouples geometry and appearance via explicit primitives and neural Gaussians for fast, high-quality streaming monocular 3D reconstruction with reduced redundancy.

Changjian Jiang, Kerui Ren, Xudong Li, Kaiwen Song and 8 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026 · ▲ 21 on Hugging Face

Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations

SCHEMA evaluates scientific agent hallucinations via topology-aware diagnostics, showing errors cluster at connected knowledge hubs and correct answers often stem from flawed reasoning.

Xinshun Feng, Ziqi Miao, Lijun Li, Jing Shao

Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

TALON: Confidence-Aware Speculative Decoding with Adaptive Token Trees

TALON is a training-free adaptive tree expansion framework for speculative decoding that allocates draft token budgets across layers to construct context-dependent deep-and-narrow or shallow-and-wide trees, achieving up to 5.16x speedup over autoregressive decoding.

Tianyu Liu, Qitan Lv, Yuhao Shen, Jun Zhang and 2 more

Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Sketch Then Paint: Hierarchical Reinforcement Learning for Diffusion Multi-Modal Large Language Models

Hierarchical Token GRPO integrates sketch-then-paint stages and hierarchical credit assignment into diffusion multi-modal LLM reinforcement learning, substantially improving image quality and benchmark scores.

Siqi Luo, Jianghan Shen, Yi Xin, Huayu Zheng and 8 more

Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Understanding Generalization through Decision Pattern Shift

Decision Pattern Shift uses GradCAM channel contributions to define generalization as stable internal decision patterns, finding DPS correlates linearly with the generalization gap and unifies failure modes.

Huiqi Deng, Yibo Li, Quanshi Zhang, Peng Zhang and 2 more

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

OmniTraffic: A Controllable Generation Pipeline and Benchmark for Spatio-Temporal Traffic Reasoning

OmniTraffic introduces a controllable 3D traffic generation pipeline and benchmark with 8M VQA samples for spatio-temporal reasoning, revealing large model gaps and improved real-world performance via simulated fine-tuning.

Maonan Wang, Zhengyan Huang, Kemou Jiang, Yuhang Fu and 12 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Synthetic Pre-Pre-Training Improves Language Model Robustness to Noisy Pre-Training Data

Synthetic pre-pre-training improves language model robustness to noisy pre-training data by inhibiting noise self-modeling and reducing required natural-text tokens by up to 49%.

Xu Guo, Runyu Peng, Jian Tong, Yunhua Zhou and 3 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

To Call or Not to Call: Diagnosing Intrinsic Over-Calling Bias in LLM Agents

LLM agents suffer intrinsic over-calling bias from an activation-independent call offset, which sparse autoencoders diagnose and steering corrects to improve accuracy.

Wei Shi, Ziheng Peng, Sihang Li, Xiting Wang and 3 more

Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026