3 citations · 3 across the 5 of their papers we have counts for
7 papers
Learning Perceptual Rules in a Fragmented Visual World
Tianqin Li, Isaac Lin, Leiran Song +1
Visual perception organizes incomplete sensory evidence into coherent structures, but where do the principles governing this organization come from? We investigate whether learning…
Witness Overlap: Directional Provenance Inside Open-Weight Model Families
Siyuan Li, Haoxuan Zeng, Xin Luo +6
Open-weight models are often released, fine-tuned, aligned, merged, and re-released, making provenance audits ask not only whether checkpoints are related, but also which checkpoin…
The Memory Curse: How Expanded Recall Erodes Cooperative Intent in LLM Agents
Jiayuan Liu, Tianqin Li, Shiyi Du +7
Context window expansion is often treated as a straightforward capability upgrade for LLMs, but we find it systematically fails in multi-agent social dilemmas. Across 7 LLMs and 4…
Learning More by Seeing Less: Structure First Learning for Efficient, Transferable, and Human-Aligned Vision
Tianqin Li, George Liu, Tai Sing Lee
Despite remarkable progress in computer vision, modern recognition systems remain fundamentally limited by their dependence on rich, redundant visual inputs. In contrast, humans ca…
From Local Cues to Global Percepts: Emergent Gestalt Organization in Self-Supervised Vision Models
Tianqin Li, Ziqi Wen, Leiran Song +3
Human vision organizes local cues into coherent global forms using Gestalt principles like closure, proximity, and figure-ground assignment -- functions reliant on global spatial s…
Emergence of Shape Bias in Convolutional Neural Networks through Activation Sparsity
Tianqin Li, Ziqi Wen, Yangfan Li +1
Current deep-learning models for object recognition are known to be heavily biased toward texture. In contrast, human visual systems are known to be biased toward shape and structu…
Does resistance to style-transfer equal Global Shape Bias? Measuring network sensitivity to global shape configuration
Ziqi Wen, Tianqin Li, Zhi Jing +1
Deep learning models are known to exhibit a strong texture bias, while human tends to rely heavily on global shape structure for object recognition. The current benchmark for evalu…