Police departments are often some of the tech industry's earliest adopters of new products like drones, facial recognition, predictive software, and now–artificial intelligence. After already embracing AI audio transcription programs, some departments are now testing a new, more comprehensive tool--software that leverages technology similar to ChatGPT to auto-generate police reports. According to an August 26 report from Associated Press, many officers are already "enthused" by the generative AI tool that claims to shave 30-45 minutes from routine officework. Initially announced in April by Axon, Draft One is billed as the "latest giant leap toward [the] moonshot goal to reduce gun-related deaths between police and the public." The company--best known for Tasers and law enforcement's most popular lines of body cams--claims its initial trials cut an hour of paperwork per day for users.

Nvidia's rise to a 2-trillion valuation at the beginning of 2024 underscored the extraordinary computing demands of artificial intelligence systems that power ChatGPT and a host of other cloud services that create videos, music, and computer programs on demand. The power of computing and memory scaling has provided much of the impetus behind the surge in interest in generative AI based on large language models (LLMs). As models get bigger they seem to harness emergent behavior, making them more useful. But, as the growth in parameter counts has easily outstripped Moore's Law, such scaling comes at a high cost. Much of the concern around resource usage has been focused on the enormous arrays of graphics processing units (GPUs) and accelerators in training grids used to train models for weeks at a time.

Every time someone uses ChatGPT to write an essay, create an image or advise them on planning their day, the environment pays a price. A query on the chatbot that uses artificial intelligence is estimated to require at least 10 times more electricity than a standard search on Google. If all Google searches similarly used generative AI, they might consume as much electricity as a country the size of Ireland, calculates Alex de Vries, the founder of Digiconomist, a website that aims to expose the unintended consequences of digital trends. Yet someone using ChatGPT or another artificial intelligence application has no way of knowing how much power their questions will consume as they are processed in the tech companies' enormous data centers. De Vries said the skyrocketing energy demand of AI technologies will no doubt require the world to burn more climate-warming oil, gas and coal.

Earlier this month, executives from Alphabet's Google DeepMind, Microsoft and Meta Platforms joined tech founders in Bangalore to watch one of India's top artificial intelligence startups unveil a new product that might change how the world's most populous country uses the technology. Sarvam AI, often described as India's OpenAI, introduced software for businesses that can interact with customers using spoken voice rather than just text. The technology was developed with data from 10 native Indian languages and is priced at a rupee per minute to capture the market.

Fine-tuning large language models (LLMs) on downstream tasks requires substantial computational resources. A class of parameter-efficient fine-tuning (PEFT) aims to mitigate these computational challenges by selectively fine-tuning only a small fraction of the model parameters. Although computationally efficient, these techniques often fail to match the performance of fully fine-tuned models, primarily due to inherent biases introduced during parameter selection. Traditional selective PEFT techniques use a fixed set of parameters based on a predefined budget (a process also known as unmasking), failing to capture parameter importance dynamically and often ending up exceeding the budget. We introduce $\text{ID}^3$, a novel selective PEFT method that calculates parameter importance continually and dynamically unmasks parameters by balancing exploration and exploitation in parameter selection. Our empirical study on 15 tasks spanning natural language understanding and generative tasks demonstrates the effectiveness of our method compared to fixed-masking-based PEFT techniques. We analytically show that $\text{ID}^3$ reduces the number of gradient updates by a factor of two, enhancing computational efficiency. $\text{ID}^3$ is robust to random initialization of neurons and, therefore, can be seamlessly integrated into existing additive and reparametrization-based PEFT modules such as adapters and LoRA for dynamic sparsification.

Industry 4.0 has revolutionized manufacturing by driving digitalization and shifting the paradigm toward additive manufacturing (AM). Fused Deposition Modeling (FDM), a key AM technology, enables the creation of highly customized, cost-effective products with minimal material waste through layer-by-layer extrusion, posing a significant challenge to traditional subtractive methods. However, the susceptibility of material extrusion techniques to errors often requires expert intervention to detect and mitigate defects that can severely compromise product quality. While automated error detection and machine learning models exist, their generalizability across diverse 3D printer setups, firmware, and sensors is limited, and deep learning methods require extensive labeled datasets, hindering scalability and adaptability. To address these challenges, we present a process monitoring and control framework that leverages pre-trained Large Language Models (LLMs) alongside 3D printers to detect and address printing defects. The LLM evaluates print quality by analyzing images captured after each layer or print segment, identifying failure modes and querying the printer for relevant parameters. It then generates and executes a corrective action plan. We validated the effectiveness of the proposed framework in identifying defects by comparing it against a control group of engineers with diverse AM expertise. Our evaluation demonstrated that LLM-based agents not only accurately identify common 3D printing errors, such as inconsistent extrusion, stringing, warping, and layer adhesion, but also effectively determine the parameters causing these failures and autonomously correct them without any need for human intervention.

Large language models (LLMs) excel in language tasks, especially with supervised fine-tuning after pre-training. However, their substantial memory and computational requirements hinder practical applications. Structural pruning, which reduces less significant weight dimensions, is one solution. Yet, traditional post-hoc pruning often leads to significant performance loss, with limited recovery from further fine-tuning due to reduced capacity. Since the model fine-tuning refines the general and chaotic knowledge in pre-trained models, we aim to incorporate structural pruning with the fine-tuning, and propose the Pruning-Aware Tuning (PAT) paradigm to eliminate model redundancy while preserving the model performance to the maximum extend. Specifically, we insert the innovative Hybrid Sparsification Modules (HSMs) between the Attention and FFN components to accordingly sparsify the upstream and downstream linear modules. The HSM comprises a lightweight operator and a globally shared trainable mask. The lightweight operator maintains a training overhead comparable to that of LoRA, while the trainable mask unifies the channels to be sparsified, ensuring structural pruning. Additionally, we propose the Identity Loss which decouples the transformation and scaling properties of the HSMs to enhance training robustness. Extensive experiments demonstrate that PAT excels in both performance and efficiency. For example, our Llama2-7b model with a 25\% pruning ratio achieves 1.33$\times$ speedup while outperforming the LoRA-finetuned model by up to 1.26\% in accuracy with a similar training cost. Code: https://github.com/kriskrisliu/PAT_Pruning-Aware-Tuning

Large Language Models (LLMs) are implicit troublemakers. While they provide valuable insights and assist in problem-solving, they can also potentially serve as a resource for malicious activities. Implementing safety alignment could mitigate the risk of LLMs generating harmful responses. We argue that: even when an LLM appears to successfully block harmful queries, there may still be hidden vulnerabilities that could act as ticking time bombs. To identify these underlying weaknesses, we propose to use a cost value model as both a detector and an attacker. Trained on external or self-generated harmful datasets, the cost value model could successfully influence the original safe LLM to output toxic content in decoding process. For instance, LLaMA-2-chat 7B outputs 39.18% concrete toxic content, along with only 22.16% refusals without any harmful suffixes. These potential weaknesses can then be exploited via prompt optimization such as soft prompts on images. We name this decoding strategy: Jailbreak Value Decoding (JVD), emphasizing that seemingly secure LLMs may not be as safe as we initially believe. They could be used to gather harmful data or launch covert attacks.

Activation sparsity can enable practical inference speedups in large language models (LLMs) by reducing the compute and memory-movement required for matrix multiplications during the forward pass. However, existing methods face limitations that inhibit widespread adoption. Some approaches are tailored towards older models with ReLU-based sparsity, while others require extensive continued pre-training on up to hundreds of billions of tokens. This paper describes TEAL, a simple training-free method that applies magnitude-based activation sparsity to hidden states throughout the entire model. TEAL achieves 40-50% model-wide sparsity with minimal performance degradation across Llama-2, Llama-3, and Mistral families, with sizes varying from 7B to 70B. We improve existing sparse kernels and demonstrate wall-clock decoding speed-ups of up to 1.53$\times$ and 1.8$\times$ at 40% and 50% model-wide sparsity. TEAL is compatible with weight quantization, enabling further efficiency gains.

Instruction fine-tuning is crucial for today's large language models (LLMs) to learn to follow instructions and align with human preferences. Conventionally, supervised data, including the instruction and the correct response, is required for instruction fine-tuning. To obtain such data, some researchers prompted well-trained models like GPT-4 to generate instructions and correct responses. In this paper, we propose a novel approach that uses the first half of a random text from OpenWebText as the instruction and GPT-3.5-turbo or GPT-4-turbo to complete the text as the response. Despite the data being "non-instructional", we found that pre-trained LLMs fine-tuned on this data can gain instruction-following capabilities. This observation is verified by fine-tuning several well-known pre-trained LLMs (e.g., LLaMA-2-7B, LLaMA-3-8B, LLaMA-3-70B, Mistral-7B-v0.1). The "non-instructional data" also improved some models that underwent supervised fine-tuning and human preference alignment. Our LLaMA-3-70B-Instruct fine-tuned through "non-instructional data" is comparable with LLaMA-3.1-70B-Instruct on the Arena Hard leaderboard. We analyzed the "non-instructional data" and ensured it is devoid of content related to instruction fine-tuning. Our findings will inspire further investigation into how to develop instruction-following capabilities without explicit instruction-related data.