Large Language Models (LLMs) are frequently updated due to data or architecture changes to improve their performance. When updating models, developers often focus on increasing overall performance metrics with less emphasis on being compatible with previous model versions. However, users often build a mental model of the functionality and capabilities of a particular machine learning model they are interacting with. They have to adapt their mental model with every update -- a draining task that can lead to user dissatisfaction. In practice, fine-tuned downstream task adapters rely on pretrained LLM base models. When these base models are updated, these user-facing downstream task models experience instance regression or negative flips -- previously correct instances are now predicted incorrectly. This happens even when the downstream task training procedures remain identical. Our work aims to provide seamless model updates to a user in two ways. First, we provide evaluation metrics for a notion of compatibility to prior model versions, specifically for generative tasks but also applicable for discriminative tasks. We observe regression and inconsistencies between different model versions on a diverse set of tasks and model updates. Second, we propose a training strategy to minimize the number of inconsistencies in model updates, involving training of a compatibility model that can enhance task fine-tuned language models. We reduce negative flips -- instances where a prior model version was correct, but a new model incorrect -- by up to 40% from Llama 1 to Llama 2.

This paper presents a novel and comprehensive solution to enhance both the robustness and efficiency of question answering (QA) systems through supervised contrastive learning (SCL). Training a high-performance QA system has become straightforward with pre-trained language models, requiring only a small amount of data and simple fine-tuning. However, despite recent advances, existing QA systems still exhibit significant deficiencies in functionality and training efficiency. We address the functionality issue by defining four key tasks: user input intent classification, out-of-domain input detection, new intent discovery, and continual learning. We then leverage a unified SCL-based representation learning method to efficiently build an intra-class compact and inter-class scattered feature space, facilitating both known intent classification and unknown intent detection and discovery. Consequently, with minimal additional tuning on downstream tasks, our approach significantly improves model efficiency and achieves new state-of-the-art performance across all tasks.

Words often carry different meanings for people from diverse backgrounds. Today's era of social polarization demands that we choose words carefully to prevent miscommunication, especially in political communication and journalism. To address this issue, we introduce the Bridging Dictionary, an interactive tool designed to illuminate how words are perceived by people with different political views. The Bridging Dictionary includes a static, printable document featuring 796 terms with summaries generated by a large language model. These summaries highlight how the terms are used distinctively by Republicans and Democrats. Additionally, the Bridging Dictionary offers an interactive interface that lets users explore selected words, visualizing their frequency, sentiment, summaries, and examples across political divides. We present a use case for journalists and emphasize the importance of human agency and trust in further enhancing this tool. The deployed version of Bridging Dictionary is available at https://dictionary.ccc-mit.org/.

Multimodal large language models (MLLMs) have become the cornerstone of today's generative AI ecosystem, sparking intense competition among tech giants and startups. In particular, an MLLM generates a text response given a prompt consisting of an image and a question. While state-of-the-art MLLMs use safety filters and alignment techniques to refuse unsafe prompts, in this work, we introduce MLLM-Refusal, the first method that induces refusals for safe prompts. In particular, our MLLM-Refusal optimizes a nearly-imperceptible refusal perturbation and adds it to an image, causing target MLLMs to likely refuse a safe prompt containing the perturbed image and a safe question. Specifically, we formulate MLLM-Refusal as a constrained optimization problem and propose an algorithm to solve it. Our method offers competitive advantages for MLLM model providers by potentially disrupting user experiences of competing MLLMs, since competing MLLM's users will receive unexpected refusals when they unwittingly use these perturbed images in their prompts. We evaluate MLLM-Refusal on four MLLMs across four datasets, demonstrating its effectiveness in causing competing MLLMs to refuse safe prompts while not affecting non-competing MLLMs. Furthermore, we explore three potential countermeasures -- adding Gaussian noise, DiffPure, and adversarial training. Our results show that they are insufficient: though they can mitigate MLLM-Refusal's effectiveness, they also sacrifice the accuracy and/or efficiency of the competing MLLM. The code is available at https://github.com/Sadcardation/MLLM-Refusal.

Language technologies are becoming more pervasive in our everyday lives, and they are also being applied in critical domains involving health, justice, and education. Given the importance of these applications and how they may affect our quality of life, it has become crucial to assess the errors they may make. In characterizing patterns of error, it has been found that systems obtained by machine-learning(ML) techniques from large quantities of text, such as large language models (LLMs), reproduce and amplify stereotypes.4 When deployed in actual applications, amplification of stereotypes can result in discriminatory behavior considered harmful in many jurisdictions. This kind of behavior is known as social bias, in that errors are distributed unevenly across social groups.

Telling a well-crafted joke that hits just right might seem like it requires a uniquely human touch. But there's bad news for comedians - as researchers from the University of Southern California say that AI is now funnier than most humans. In their study, ChatGPT was able to craft punchlines that were rated funnier than human efforts 70 per cent of the time. And this isn't a laughing matter, as the researchers warn that joke-writing robots could pose a'serious employment threat' to professional comedians. So, can you tell which of these jokes were written by a robot?

Foundation models, e.g., large language models (LLMs), trained on internet-scale data possess zero-shot generalization capabilities that make them a promising technology towards detecting and mitigating out-of-distribution failure modes of robotic systems. Fully realizing this promise, however, poses two challenges: (i) mitigating the considerable computational expense of these models such that they may be applied online, and (ii) incorporating their judgement regarding potential anomalies into a safe control framework. In this work, we present a two-stage reasoning framework: First is a fast binary anomaly classifier that analyzes observations in an LLM embedding space, which may then trigger a slower fallback selection stage that utilizes the reasoning capabilities of generative LLMs. These stages correspond to branch points in a model predictive control strategy that maintains the joint feasibility of continuing along various fallback plans to account for the slow reasoner's latency as soon as an anomaly is detected, thus ensuring safety. We show that our fast anomaly classifier outperforms autoregressive reasoning with state-of-the-art GPT models, even when instantiated with relatively small language models. This enables our runtime monitor to improve the trustworthiness of dynamic robotic systems, such as quadrotors or autonomous vehicles, under resource and time constraints. Videos illustrating our approach in both simulation and real-world experiments are available on this project page: https://sites.google.com/view/aesop-llm.

This paper introduces Lite-SAM, an efficient end-to-end solution for the SegEvery task designed to reduce computational costs and redundancy. Lite-SAM is composed of four main components: a streamlined CNN-Transformer hybrid encoder (LiteViT), an automated prompt proposal network (AutoPPN), a traditional prompt encoder, and a mask decoder. All these components are integrated within the SAM framework. Our LiteViT, a high-performance lightweight backbone network, has only 1.16M parameters, which is a 23% reduction compared to the lightest existing backbone network Shufflenet. We also introduce AutoPPN, an innovative end-to-end method for prompt boxes and points generation. This is an improvement over traditional grid search sampling methods, and its unique design allows for easy integration into any SAM series algorithm, extending its usability. we have thoroughly benchmarked Lite-SAM across a plethora of both public and private datasets. The evaluation encompassed a broad spectrum of universal metrics, including the number of parameters, SegEvery execution time, and accuracy. The findings reveal that Lite-SAM, operating with a lean 4.2M parameters, significantly outpaces its counterparts, demonstrating performance improvements of 43x, 31x, 20x, 21x, and 1.6x over SAM, MobileSAM, Edge-SAM, EfficientViT-SAM, and MobileSAM-v2 respectively, all the while maintaining competitive accuracy. This underscores Lite-SAM's prowess in achieving an optimal equilibrium between performance and precision, thereby setting a new state-of-the-art(SOTA) benchmark in the domain.

Natural language interfaces have exhibited considerable potential in the automation of Verilog generation derived from high-level specifications through the utilization of large language models, garnering significant attention. Nevertheless, this paper elucidates that visual representations contribute essential contextual information critical to design intent for hardware architectures possessing spatial complexity, potentially surpassing the efficacy of natural-language-only inputs. Expanding upon this premise, our paper introduces an open-source benchmark for multi-modal generative models tailored for Verilog synthesis from visual-linguistic inputs, addressing both singular and complex modules. Additionally, we introduce an open-source visual and natural language Verilog query language framework to facilitate efficient and user-friendly multi-modal queries. To evaluate the performance of the proposed multi-modal hardware generative AI in Verilog generation tasks, we compare it with a popular method that relies solely on natural language. Our results demonstrate a significant accuracy improvement in the multi-modal generated Verilog compared to queries based solely on natural language. We hope to reveal a new approach to hardware design in the large-hardware-design-model era, thereby fostering a more diversified and productive approach to hardware design.

Mechanistic interpretability work attempts to reverse engineer the learned algorithms present inside neural networks. One focus of this work has been to discover 'circuits' -- subgraphs of the full model that explain behaviour on specific tasks. But how do we measure the performance of such circuits? Prior work has attempted to measure circuit 'faithfulness' -- the degree to which the circuit replicates the performance of the full model. In this work, we survey many considerations for designing experiments that measure circuit faithfulness by ablating portions of the model's computation. Concerningly, we find existing methods are highly sensitive to seemingly insignificant changes in the ablation methodology. We conclude that existing circuit faithfulness scores reflect both the methodological choices of researchers as well as the actual components of the circuit - the task a circuit is required to perform depends on the ablation used to test it. The ultimate goal of mechanistic interpretability work is to understand neural networks, so we emphasize the need for more clarity in the precise claims being made about circuits. We open source a library at https://github.com/UFO-101/auto-circuit that includes highly efficient implementations of a wide range of ablation methodologies and circuit discovery algorithms.