Topic
attention
ITNet: A Learnable Integral Transform That Unifies Convolution, Attention, and Recurrence in One Architecture
Researchers introduce ITNet, a neural architecture built on a learnable integral transform that subsumes convolution, self-attention, and recurrence as special cases. A single ITNet with a shared operator matches or exceeds specialized models on ImageNet-1K, GLUE, ModelNet40, VQA v2, and NLVR2, enabled by efficient techniques such as tiled kernel fusion and Monte Carlo integration.
Attention, Not Model Scale, Drives Human-AI Alignment in Multimodal Language Prediction, Research Finds
A study comparing five vision-language models with 600 human participants found that adding visual context significantly improved human-AI alignment in language prediction, with attention maps explaining up to 70% of inter-participant variance. The research indicates that attention to informative cues, not model scale, is the primary driver of alignment.
Attention as Coupling: New Fast-Slow ODE Framework Aims to Improve Transformer Efficiency
A new research paper proposes a fast-slow ordinary differential equation (ODE) framework for hierarchical pretraining in transformers. The authors instantiate a neural network with a fast causal attention path and a slower pooled attention path, proving a theoretical link to stationary distributions. Empirical results at 500k tokens show neutral coupling, with wall-clock cost comparable to dense baseline.
GEASS: Gated Evidence-Adaptive Selective Caption Trust Tackles VLM Hallucination
Vision-language models often hallucinate objects, and feeding them their own captions can actually worsen accuracy. Researchers propose GEASS, a gated evidence-adaptive module that decides per query how much of the caption to trust, improving accuracy across four VLMs on two benchmarks without training or additional parameters.
LLM Agents Look at Correct Tools but Still Pick Wrong, Research Reveals Readout as Failure Point
Research by Shiyang Chen reveals that LLM agents mis-call tools not because they fail to see the right tool, but because the decision readout fails. The model attends to the correct tool 80% of the time, yet picks wrong. Readout-side interventions recover 59-91% of failures, while input-side fixes recover ≤23%.
New VeriAttn Technique Accelerates Verifiable LLM Inference on TEE-GPU Systems
Researchers propose VeriAttn, a communication-efficient TEE-GPU attention mechanism for verifiable LLM inference. By offloading attention computations to the GPU while the TEE performs verification, VeriAttn achieves 2.60-3.38x acceleration for prefill and 3.86-5.42x for decoding over the TSDP baseline on Intel TDX.