Xiaomi (China)

For industrial-scale text-to-SQL, supplying the entire database schema to Large Language Models (LLMs) is impractical due to context window limits and irrelevant noise. Schema linking, which filters the schema to a relevant subset, is therefore critical. However, existing methods incur prohibitive costs, struggle to trade off recall and noise, and scale poorly to large databases. We present AutoLink, an autonomous agent framework that reformulates schema linking as an iterative, agent-driven process. Guided by an LLM, AutoLink dynamically explores and expands the linked schema subset, progressively identifying necessary schema components without inputting the full database schema. Our experiments demonstrate AutoLink's superior performance, achieving state-of-the-art strict schema linking recall of 97.4% on Bird-Dev and 91.2% on Spider 2.0-Lite, with competitive execution accuracy, i.e., 68.7% EX on Bird-Dev (better than CHESS) and 34.9% EX on Spider 2.0-Lite (ranking 2nd on the official leaderboard). Crucially, AutoLink exhibits exceptional scalability, maintaining high recall, efficient token consumption, and robust execution accuracy on large schemas (e.g., over 3,000 columns) where existing methods severely degrade—making it a highly scalable, high-recall schema-linking solution for industrial text-to-SQL systems.

Conventional FP-CIMs suffer from fixed preserved bit-width (PBW), limiting their adaptability and efficiency. This work proposes the first MXFP-CIM macro enabling wide-range adaptive PBW, featuring: (1) A serial dual-bit-sliding scheme; (2) A harmless data mapping scheme with a hierarchical hidden-bit decoder; (3) An adjustable-PBW MXFP-MAC circuit via twin-stage allocation. The 28nm MXFP-CIM macro achieves a peak energy efficiency of 127.54TFLOPS/W in the MXFP6/6 mode.

As autonomous vehicle technology continues to evolve, ensuring its safety in complex and dynamic environments has become a critical challenge. Knowledge Graphs (KGs) and Large Language Models (LLMs), as two cutting-edge approaches representing the for

The direction of target speech is a crucial clue for multichannel speech enhancement. But its real-time estimation is usually very challenging, especially when the source is moving and in far-field reverb environments. Instead of directly estimating the direction, we propose a direction-guided spatial attention to automatically focus on the target source and incorporate it into masking neural beamformers for speech enhancement. Neural beamformers are trained to learn to preserve signals from the look direction areas and suppress signals from other direction areas as much as possible. Spatial attention is guided by an extra classification objective of the target direction to focus on the target direction area and select an appropriate neural beamformer to listen to the direction area of target speech. Systematic experiments demonstrate that the proposed approach improves the performances of both speech enhancement and far-field speech recognition against prior methods. <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup><sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup>Corresponding Author: Yaran Chen

Organic light-emitting diode (OLED) displays have been widely used in smartphones in recent years. Manufacturers use the stroboscopic visibility measure (SVM) to characterize and compare the performance of OLED displays using the pulse-width modulation (PWM) dimming method in producing perceivable flicker, although the SVM was developed for lighting products. This study investigated how the dimming frequency and display luminance of an OLED smartphone display affected the stroboscopic effect by presenting an oscillating arm in front of the display. The results showed that the stroboscopic effect was visible when the dimming frequency was 480 and 960Hz with the display luminance below 40cd/m2, and when the dimming frequency was 1440Hz, with the display luminance at 5cd/m2, the threshold of the duty cycle was around 50%. When the dimming frequency was 1920 or 2160Hz, the stroboscopic effect was invisible. The results also suggested that for OLED displays, the SVM may still be used, but an SVM of 1.5 to 2 can be used as a conservative threshold instead under the experimental conditions.

Large Language Models (LLMs), built on Transformer architectures, exhibit remarkable generalization across a wide range of tasks. However, fine-tuning these models for specific tasks remains resource-intensive due to their extensive parameterization. In this paper, we explore two remarkable phenomena related to the attention mechanism during the fine-tuning of LLMs (where Wq, Wk, and Wv denote the weights of the query, key, and value layers, respectively). The first phenomenon, termed “Unequal Importance of Attention Matrices”, highlights the impact of fine-tuning different weight matrices. It shows that optimizing the Wv matrix yields significantly better performance than optimizing the Wk matrix. Fine-tuning only the Wq and Wv matrices is computationally efficient while delivering results comparable to, or even better than fine-tuning all three matrices (Wq, Wk, and Wv). The second phenomenon, “Attention Matrices with Customized Learning Rate Lead to Better Convergence”, emphasizes the importance of assigning distinct learning rates to these matrices. Specifically, a higher learning rate for the Wv matrix compared to Wq and Wk accelerates convergence and improves performance. Building on these insights, we propose a new strategy that improves fine-tuning efficiency in terms of both storage and time. Experimental results on benchmark datasets validate the effectiveness of this approach, supporting our theoretical findings. Our analysis lays the theoretical groundwork for configuring and improving algorithms in LLMs fine-tuning.