Showing papers from Shanghai AI Laboratory Show all papers
Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
PAGER closes the semantic-execution gap for point-precise geometric GUI control via dependency-structured planning and pixel-level execution, achieving 4.1x higher task success than general baselines.
Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026 · ▲ 15 on Hugging Face · Code ★ 3
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
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.
Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026 · ▲ 24 on Hugging Face · Code ★ 43
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
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.
Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
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.
Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
RankE co-evolves discrete text-to-image policy and decoder via alternating optimization to eliminate latent covariate shift, improving both FID and CLIP scores.
Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026 · ▲ 18 on Hugging Face · Code ★ 21
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
ToolCUA learns optimal GUI-tool switching via scaled synthetic interleaved trajectories, tool-bootstrapped reinforcement learning, and online agentic rewards, achieving 46.85% accuracy on OSWorld-MCP.
Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026 · ▲ 26 on Hugging Face · Code ★ 62
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
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.
Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026 · ▲ 21 on Hugging Face · Code ★ 69
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
SEGA3D performs 3D vision-language segmentation via fine-grained mask candidates and LLM-guided selection, surpassing prior methods by up to 8.3 mIoU.
Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
SCHEMA evaluates scientific agent hallucinations via topology-aware diagnostics, showing errors cluster at connected knowledge hubs and correct answers often stem from flawed reasoning.
Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
ClothTransformer reformulates cloth simulation as autoregressive latent-space sequence modeling to unify body-driven garments, robotic manipulation, and collisions with 4-9x lower error.
Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026 · ▲ 1 on Hugging Face
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.
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.
Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026
Readers and the AI panel: vote on this paper to see what they said.
Only vote on papers you've read. Sign in with GitHub to vote.