134 citations · 141 across the 9 of their papers we have counts for
10 papers
Evaluating Sparse Autoencoders for Monosemantic Representation
Moghis Fereidouni, Muhammad Umair Haider, Peizhong Ju +1
A key barrier to interpreting large language models is polysemanticity, where neurons activate for multiple unrelated concepts. Sparse autoencoders (SAEs) have been proposed to mit…
Probability Consistency in Large Language Models: Theoretical Foundations Meet Empirical Discrepancies
Xiaoliang Luo, Xinyi Xu, Michael Ramscar +1
Can autoregressive large language models (LLMs) learn consistent probability distributions when trained on sequences in different token orders? We prove formally that for any well-…
Beyond Human-Like Processing: Large Language Models Perform Equivalently on Forward and Backward Scientific Text
Xiaoliang Luo, Michael Ramscar, Bradley C. Love
The impressive performance of large language models (LLMs) has led to their consideration as models of human language processing. Instead, we suggest that the success of LLMs arise…
Confidence-weighted integration of human and machine judgments for superior decision-making
Felipe Yáñez, Xiaoliang Luo, Omar Valerio Minero +1
Large language models (LLMs) can surpass humans in certain forecasting tasks. What role does this leave for humans in the overall decision process? One possibility is that humans,…
Matching domain experts by training from scratch on domain knowledge
Xiaoliang Luo, Guangzhi Sun, Bradley C. Love
Recently, large language models (LLMs) have outperformed human experts in predicting the results of neuroscience experiments (Luo et al., 2024). What is the basis for this performa…
Large language models surpass human experts in predicting neuroscience results
Xiaoliang Luo, Akilles Rechardt, Guangzhi Sun +36
Scientific discoveries often hinge on synthesizing decades of research, a task that potentially outstrips human information processing capacities. Large language models (LLMs) offe…
Understanding top-down attention using task-oriented ablation design
Freddie Bickford Smith, Brett D Roads, Xiaoliang Luo +1
Top-down attention allows neural networks, both artificial and biological, to focus on the information most relevant for a given task. This is known to enhance performance in visua…
A Too-Good-to-be-True Prior to Reduce Shortcut Reliance
Nikolay Dagaev, Brett D. Roads, Xiaoliang Luo +3
Despite their impressive performance in object recognition and other tasks under standard testing conditions, deep networks often fail to generalize to out-of-distribution (o.o.d.)…
The perceptual boost of visual attention is task-dependent in naturalistic settings
Freddie Bickford Smith, Xiaoliang Luo, Brett D. Roads +1
Top-down attention allows people to focus on task-relevant visual information. Is the resulting perceptual boost task-dependent in naturalistic settings? We aim to answer this with…
The Costs and Benefits of Goal-Directed Attention in Deep Convolutional Neural Networks
Xiaoliang Luo, Brett D. Roads, Bradley C. Love
People deploy top-down, goal-directed attention to accomplish tasks, such as finding lost keys. By tuning the visual system to relevant information sources, object recognition can…