相关论文: Latent Introspection: Models Can Detect Prior Conc…

Backdoor attacks are an insidious security threat against machine learning models. Adversaries can manipulate the predictions of compromised models by inserting triggers into the training phase. Various backdoor attacks have been devised…

Concept Bottleneck Models (CBMs) aim to enhance interpretability by structuring predictions around human-understandable concepts. However, unintended information leakage, where predictive signals bypass the concept bottleneck, compromises…

Self-consciousness, the introspection of one's existence and thoughts, represents a high-level cognitive process. As language models advance at an unprecedented pace, a critical question arises: Are these models becoming self-conscious?…

In parallel to their overwhelming success across NLP tasks, language ability of deep Transformer networks, pretrained via language modeling (LM) objectives has undergone extensive scrutiny. While probing revealed that these models encode a…

Generation of photo-realistic images, semantic editing and representation learning are a few of many potential applications of high resolution generative models. Recent progress in GANs have established them as an excellent choice for such…

To plan safe maneuvers and act with foresight, autonomous vehicles must be capable of accurately predicting the uncertain future. In the context of autonomous driving, deep neural networks have been successfully applied to learning…

Knowledge distillation from large language models (LLMs) assumes that the teacher's output distribution is a high-quality training signal. On reasoning tasks, this assumption is frequently violated. A model's intermediate representations…

Purpose: Our study presents an enhanced approach to medical image caption generation by integrating concept detection into attention mechanisms. Method: This method utilizes sophisticated models to identify critical concepts within medical…

Understanding the internal thinking process of Large Language Models (LLMs) and the cause of hallucinations remains a key challenge. To this end, we introduce latent debate, a novel framework for interpreting model predictions through the…

Concept Bottleneck Models (CBM) are inherently interpretable models that factor model decisions into human-readable concepts. They allow people to easily understand why a model is failing, a critical feature for high-stakes applications.…

We introduce a simple, yet novel entropy-based framework to drive token efficiency in large language models during reasoning tasks. Our approach uses Shannon entropy from token-level logprobs as a confidence signal to enable early stopping,…

The possibility of LLM self-awareness and even sentience is gaining increasing public attention and has major safety and policy implications, but the science of measuring them is still in a nascent state. Here we introduce a novel…

Architectural obfuscation - e.g., permuting hidden-state tensors, linearly transforming embedding tables, or remapping tokens - has recently gained traction as a lightweight substitute for heavyweight cryptography in privacy-preserving…

Latent visual reasoning aims to mimic human's imagination process by meditating through hidden states of Multimodal Large Language Models. While recognized as a promising paradigm for visual reasoning, the underlying mechanisms driving its…

Intuitive observations show that a baby may inherently possess the capability of recognizing a new visual concept (e.g., chair, dog) by learning from only very few positive instances taught by parent(s) or others, and this recognition…

In contrast to RNNs, which compress their history into a single hidden state, Transformers can attend to all past tokens directly. However, standard Transformers rely solely on the hidden state from the previous layer to represent the…

This paper investigates the integration of the Learning Using Privileged Information (LUPI) paradigm in object detection to exploit fine-grained, descriptive information available during training but not at inference. We introduce a…

Recent latent-space monitoring techniques have shown promise as defenses against LLM attacks. These defenses act as scanners that seek to detect harmful activations before they lead to undesirable actions. This prompts the question: Can…

Pre-trained language models demonstrate general intelligence and common sense, but long inputs quickly become a bottleneck for memorizing information at inference time. We resurface a simple method, Memorizing Transformers (Wu et al.,…

Concept Bottleneck Models (CBMs) first map raw input(s) to a vector of human-defined concepts, before using this vector to predict a final classification. We might therefore expect CBMs capable of predicting concepts based on distinct…