Language model continual learning (CL) has recently attracted significant interest for its ability to adapt large language models (LLMs) to dynamic real-world scenarios without retraining. A major challenge in this domain is catastrophic forgetting, where models lose previously acquired knowledge upon learning new tasks. Existing approaches commonly utilize multiple parameter-efficient fine-tuning (PEFT) blocks to acquire task-specific knowledge, yet these methods are inefficient and fail to leverage potential knowledge transfer across tasks. In this paper, we introduce a novel CL framework for language models, named Task Skill Localization and Consolidation (TaSL), which boosts knowledge transfer without depending on memory replay. TaSL initially segregates the model into 'skill units' based on parameter dependencies, allowing for more precise control. Subsequently, it employs a novel group-wise skill localization technique to ascertain the importance distribution of skill units for a new task. By comparing this importance distribution with those from previous tasks, we implement a fine-grained skill consolidation strategy that retains task-specific knowledge, thereby preventing forgetting, and updates task-shared knowledge, which facilitates bi-directional knowledge transfer. As a result, TaSL achieves an optimal balance between retaining prior knowledge and excelling in new tasks. TaSL also demonstrates strong generalizability, making it suitable for various base models and adaptable to PEFT methods like LoRA. Furthermore, it offers notable extensibility, supporting enhancements through integration with memory replay techniques. Comprehensive experiments conducted on two CL benchmarks, involving models ranging from 220M to 7B parameters, affirm the effectiveness of TaSL and its variants across different settings.

Software security remains a critical concern, particularly as junior developers, often lacking comprehensive knowledge of security practices, contribute to codebases. While there are tools to help developers proactively write secure code, their actual effectiveness in helping developers fix their vulnerable code remains largely unmeasured. Moreover, these approaches typically focus on classifying and localizing vulnerabilities without highlighting the specific code segments that are the root cause of the issues, a crucial aspect for developers seeking to fix their vulnerable code. To address these challenges, we conducted a comprehensive study evaluating the efficacy of existing methods in helping junior developers secure their code. Our findings across five types of security vulnerabilities revealed that current tools enabled developers to secure only 36.2\% of vulnerable code. Questionnaire results from these participants further indicated that not knowing the code that was the root cause of the vulnerability was one of their primary challenges in repairing the vulnerable code. Informed by these insights, we developed an automated vulnerability root cause (RC) toolkit called T5-RCGCN, that combines T5 language model embeddings with a graph convolutional network (GCN) for vulnerability classification and localization. Additionally, we integrated DeepLiftSHAP to identify the code segments that were the root cause of the vulnerability. We tested T5-RCGCN with 56 junior developers across three datasets, showing a 28.9\% improvement in code security compared to previous methods. Developers using the tool also gained a deeper understanding of vulnerability root causes, resulting in a 17.0\% improvement in their ability to secure code independently. These results demonstrate the tool's potential for both immediate security enhancement and long-term developer skill growth.

We apply a state-of-the-art difference-in-differences approach to estimate the impact of ChatGPT's release on the writing style of condensed matter papers on arXiv. Our analysis reveals a statistically significant improvement in the English quality of abstracts written by non-native English speakers. Importantly, this improvement remains robust even after accounting for other potential factors, confirming that it can be attributed to the release of ChatGPT. This indicates widespread adoption of the tool. Following the release of ChatGPT, there is a significant increase in the use of unique words, while the frequency of rare words decreases. Across language families, the changes in writing style are significant for authors from the Latin and Ural-Altaic groups, but not for those from the Germanic or other Indo-European groups.

The rapid development of large language models (LLMs) has spurred extensive research into their domain-specific capabilities, particularly mathematical reasoning. However, most open-source LLMs focus solely on mathematical reasoning, neglecting the integration with visual injection, despite the fact that many mathematical tasks rely on visual inputs such as geometric diagrams, charts, and function plots. To fill this gap, we introduce \textbf{MultiMath-7B}, a multimodal large language model that bridges the gap between math and vision. \textbf{MultiMath-7B} is trained through a four-stage process, focusing on vision-language alignment, visual and math instruction-tuning, and process-supervised reinforcement learning. We also construct a novel, diverse and comprehensive multimodal mathematical dataset, \textbf{MultiMath-300K}, which spans K-12 levels with image captions and step-wise solutions. MultiMath-7B achieves state-of-the-art (SOTA) performance among open-source models on existing multimodal mathematical benchmarks and also excels on text-only mathematical benchmarks. Our model and dataset are available at {\textcolor{blue}{\url{https://github.com/pengshuai-rin/MultiMath}}}.

Cosine similarity between two documents can be computed using token embeddings formed by Large Language Models (LLMs) such as GPT-4, and used to categorize those documents across a range of uses. However, these similarities are ultimately dependent on the corpora used to train these LLMs, and may not reflect subjective similarity of individuals or how their biases and constraints impact similarity metrics. This lack of cognitively-aware personalization of similarity metrics can be particularly problematic in educational and recommendation settings where there is a limited number of individual judgements of category or preference, and biases can be particularly relevant. To address this, we rely on an integration of an Instance-Based Learning (IBL) cognitive model with LLM embeddings to develop the Instance-Based Individualized Similarity (IBIS) metric. This similarity metric is beneficial in that it takes into account individual biases and constraints in a manner that is grounded in the cognitive mechanisms of decision making. To evaluate the IBIS metric, we also introduce a dataset of human categorizations of emails as being either dangerous (phishing) or safe (ham). This dataset is used to demonstrate the benefits of leveraging a cognitive model to measure the subjective similarity of human participants in an educational setting.

When news broke that AI-generated nude pictures of students were popping up at a Beverly Hills Middle School in February, many district officials and parents were horrified. But others said no one should have been blindsided by the spread of AI-powered "undressing" programs. "The only thing shocking about this story," one Carlsbad parent said his 14-year-old told him, "is that people are shocked." Now, a newly released report by Thorn, a tech company that works to stop the spread of child sexual abuse material, shows how common deepfake abuse has become. The proliferation coincides with the wide availability of cheap "undressing" apps and other easy-to-use, AI-powered programs to create deepfake nudes.

Language models have demonstrated remarkable performance in solving reasoning tasks; however, even the strongest models still occasionally make reasoning mistakes. Recently, there has been active research aimed at improving reasoning accuracy, particularly by using pretrained language models to "self-correct" their mistakes via multi-round prompting. In this paper, we follow this line of work but focus on understanding the usefulness of incorporating "error-correction" data directly into the pretraining stage. This data consists of erroneous solution steps immediately followed by their corrections. Using a synthetic math dataset, we show promising results: this type of pretrain data can help language models achieve higher reasoning accuracy directly (i.e., through simple auto-regression, without multi-round prompting) compared to pretraining on the same amount of error-free data. We also delve into many details, such as (1) how this approach differs from beam search, (2) how such data can be prepared, (3) whether masking is needed on the erroneous tokens, (4) the amount of error required, (5) whether such data can be deferred to the fine-tuning stage, and many others.

Training on high-quality synthetic data from strong language models (LMs) is a common strategy to improve the reasoning performance of LMs. In this work, we revisit whether this strategy is compute-optimal under a fixed inference budget (e.g., FLOPs). To do so, we investigate the trade-offs between generating synthetic data using a stronger but more expensive (SE) model versus a weaker but cheaper (WC) model. We evaluate the generated data across three key metrics: coverage, diversity, and false positive rate, and show that the data from WC models may have higher coverage and diversity, but also exhibit higher false positive rates. We then finetune LMs on data from SE and WC models in different settings: knowledge distillation, self-improvement, and a novel weak-to-strong improvement setup where a weaker LM teaches reasoning to a stronger LM. Our findings reveal that models finetuned on WC-generated data consistently outperform those trained on SE-generated data across multiple benchmarks and multiple choices of WC and SE models. These results challenge the prevailing practice of relying on SE models for synthetic data generation, suggesting that WC may be the compute-optimal approach for training advanced LM reasoners.

For d/Deaf and hard of hearing (DHH) people, captioning is an essential accessibility tool. Significant developments in artificial intelligence (AI) mean that Automatic Speech Recognition (ASR) is now a part of many popular applications. This makes creating captions easy and broadly available - but transcription needs high levels of accuracy to be accessible. Scientific publications and industry report very low error rates, claiming AI has reached human parity or even outperforms manual transcription. At the same time the DHH community reports serious issues with the accuracy and reliability of ASR. There seems to be a mismatch between technical innovations and the real-life experience for people who depend on transcription. Independent and comprehensive data is needed to capture the state of ASR. We measured the performance of eleven common ASR services with recordings of Higher Education lectures. We evaluated the influence of technical conditions like streaming, the use of vocabularies, and differences between languages. Our results show that accuracy ranges widely between vendors and for the individual audio samples. We also measured a significant lower quality for streaming ASR, which is used for live events. Our study shows that despite the recent improvements of ASR, common services lack reliability in accuracy.

In this provocation, we discuss the English dominance of the AI research community, arguing that the requirement for English language publishing upholds and reinforces broader regimes of extraction in AI. While large language models and machine translation have been celebrated as a way to break down barriers, we regard their use as a symptom of linguistic exclusion of scientists and potential readers. We propose alternative futures for a healthier publishing culture, organized around three themes: administering conferences in the languages of the country in which they are held, instructing peer reviewers not to adjudicate the language appropriateness of papers, and offering opportunities to publish and present in multiple languages. We welcome new translations of this piece. Please contact the authors if you would like to contribute one.