Accurate and verifiable large language model (LLM) simulations of human research subjects promise an accessible data source for understanding human behavior and training new AI systems. However, results to date have been limited, and few social scientists have adopted this method. In this position paper, we argue that the promise of LLM social simulations can be achieved by addressing five tractable challenges. We ground our argument in a review of empirical comparisons between LLMs and human research subjects, commentaries on the topic, and related work. We identify promising directions, including context-rich prompting and fine-tuning with social science datasets. We believe that LLM social simulations can already be used for pilot and exploratory studies, and more widespread use may soon be possible with rapidly advancing LLM capabilities. Researchers should prioritize developing conceptual models and iterative evaluations to make the best use of new AI systems.

Search-augmented language models combine web search with Large Language Models (LLMs) to improve response groundedness and freshness. However, analyzing these systems remains challenging: existing datasets are limited in scale and narrow in scope, often constrained to static, single-turn, fact-checking questions. In this work, we introduce Search Arena, a crowd-sourced, large-scale, human-preference dataset of over 24,000 paired multi-turn user interactions with search-augmented LLMs. The dataset spans diverse intents and languages, and contains full system traces with around 12,000 human preference votes. Our analysis reveals that user preferences are influenced by the number of citations, even when the cited content does not directly support the attributed claims, uncovering a gap between perceived and actual credibility. Furthermore, user preferences vary across cited sources, revealing that community-driven platforms are generally preferred and static encyclopedic sources are not always appropriate and reliable. To assess performance across different settings, we conduct cross-arena analyses by testing search-augmented LLMs in a general-purpose chat environment and conventional LLMs in search-intensive settings. We find that web search does not degrade and may even improve performance in non-search settings; however, the quality in search settings is significantly affected if solely relying on the model's parametric knowledge. We open-sourced the dataset to support future research in this direction. Our dataset and code are available at: https://github.com/lmarena/search-arena.

Large language models (LLMs) are typically trained on enormous quantities of unlicensed text, a practice that has led to scrutiny due to possible intellectual property infringement and ethical concerns. Training LLMs on openly licensed text presents a first step towards addressing these issues, but prior data collection efforts have yielded datasets too small or low-quality to produce performant LLMs. To address this gap, we collect, curate, and release the Common Pile v0.1, an eight terabyte collection of openly licensed text designed for LLM pretraining. The Common Pile comprises content from 30 sources that span diverse domains including research papers, code, books, encyclopedias, educational materials, audio transcripts, and more. Crucially, we validate our efforts by training two 7 billion parameter LLMs on text from the Common Pile: Comma v0.1-1T and Comma v0.1-2T, trained on 1 and 2 trillion tokens respectively. Both models attain competitive performance to LLMs trained on unlicensed text with similar computational budgets, such as Llama 1 and 2 7B. In addition to releasing the Common Pile v0.1 itself, we also release the code used in its creation as well as the training mixture and checkpoints for the Comma v0.1 models.

An effective approach to the development of ASR systems for low-resource languages is to fine-tune an existing multilingual end-to-end model. When the original model has been trained on large quantities of data from many languages, fine-tuning can be effective with limited training data, even when the language in question was not present in the original training data. The fine-tuning approach has been encouraged by the availability of public-domain E2E models and is widely believed to lead to state-of-the-art results. This paper, however, challenges that belief. We show that an approach combining hybrid HMMs with self-supervised models can yield substantially better performance with limited training data. This combination allows better utilisation of all available speech and text data through continued self-supervised pre-training and semi-supervised training. We benchmark our approach on Scottish Gaelic, achieving WER reductions of 32% relative over our best fine-tuned Whisper model.

AI music generation has advanced rapidly, with models like diffusion and autoregressive algorithms enabling high-fidelity outputs. These tools can alter styles, mix instruments, or isolate them. Since sound can be visualized as spectrograms, image-generation algorithms can be applied to generate novel music. However, these algorithms are typically trained on fixed datasets, which makes it challenging for them to interpret and respond to user input accurately. This is especially problematic because music is highly subjective and requires a level of personalization that image generation does not provide. In this work, we propose a human-computation approach to gradually improve the performance of these algorithms based on user interactions. The human-computation element involves aggregating and selecting user ratings to use as the loss function for fine-tuning the model. We employ a genetic algorithm that incorporates user feedback to enhance the baseline performance of a model initially trained on a fixed dataset. The effectiveness of this approach is measured by the average increase in user ratings with each iteration. In the pilot test, the first iteration showed an average rating increase of 0.2 compared to the baseline. The second iteration further improved upon this, achieving an additional increase of 0.39 over the first iteration.

--The widespread dissemination of fake news on social media has significantly impacted society, resulting in serious consequences. Conventional deep learning methodologies employing small language models (SLMs) suffer from extensive supervised training requirements and difficulties adapting to evolving news environments due to data scarcity and distribution shifts. EFND) framework to address these challenges. We further introduce a lifelong knowledge editing module based on a Mixture-of-Experts architecture to incrementally update LLMs and a replay-based continue learning method to ensure SLMs retain prior knowledge without retraining entirely. EFND significantly outperforms existed methods, effectively improving detection accuracy and adaptability in continuous emergent fake news scenarios. HE rampant spread of fake news on the Internet has already caused significant societal impact [1]. For instance, the spread of fake news during the Covid-19 pandemic has led to harmful consequences such as drug misuse and incorrect treatment methods [2]. As illustrated in Figure 2(a), fake news on emergent events evolves continuously, presenting a challenge for real-time detection systems to keep pace with its evolution. Furthermore, an alarming pattern known as "rumor resurgence" frequently occurs in social media, wherein past misinformation reappears, perpetuating its societal impact [3]. Chaozhuo Li is with School of Cyber Science and Technology, Beijing University of Posts and Telecommunications, Beijing 100876, China (e-mail: lichaozhuo@bupt.edu.cn).

In-Context Learning (ICL) empowers Large Language Models (LLMs) for rapid task adaptation without Fine-Tuning (FT), but its reliance on demonstration selection remains a critical challenge. While many-shot ICL shows promising performance through scaled demonstrations, the selection method for many-shot demonstrations remains limited to random selection in existing work. Since the conventional instance-level retrieval is not suitable for many-shot scenarios, we hypothesize that the data requirements for in-context learning and fine-tuning are analogous. To this end, we introduce a novel gradient matching approach that selects demonstrations by aligning fine-tuning gradients between the entire training set of the target task and the selected examples, so as to approach the learning effect on the entire training set within the selected examples. Through gradient matching on relatively small models, e.g., Qwen2.5-3B or Llama3-8B, our method consistently outperforms random selection on larger LLMs from 4-shot to 128-shot scenarios across 9 diverse datasets. For instance, it surpasses random selection by 4% on Qwen2.5-72B and Llama3-70B, and by around 2% on 5 closed-source LLMs. This work unlocks more reliable and effective many-shot ICL, paving the way for its broader application.

It's been instructive to see "Ballerina," which opens this week, so soon after the new "Mission: Impossible" installment. In the latter, it's hard to top Tom Cruise's intrepid stunt work, which reaches its zenith in a pair of extended sequences (one in a submarine, the other on biplanes), but the story, involving a diabolical scheme using A.I. to commandeer and launch the world's nuclear weaponry, is a mere pretext. Going to "Mission: Impossible" for the story is like going to Casablanca for the waters. In contrast, "Ballerina"--like the four John Wick films that it's spun off from--is, strangely, far better at story than at action. The first John Wick film is the weakest, because the framework for the franchise was still unformed: a retired hit man (Keanu Reeves) gets back into action to respond to a mobster's attacks.

Weekly quiz: What did Taylor Swift buy back? This week saw Ukraine mount an audacious drone attack on Russian airfields, Donald Trump ban people in 12 countries from travelling to the US, while Billie Piper returned to Doctor Who. But how much attention did you pay to what else happened in the world? Try last week's quiz, or have a go at something from the archives.Taylor SwiftMusicRelated The 43-year-old global superstar performed the first night of her six show residency in London.2 The star says he is focusing on acting after parting ways with his record company.3 hrs agoCulture8 hrs ago The annual event kicks off on Thursday and runs until the end of June.8 hrs ago Northern Ireland13 hrs ago Cally Rhodes from Shrewsbury says being selected to perform with Murs at Ludlow Castle is "unreal".13 Paul Weller cover version'the ultimate tribute' to Friel The family Eamon Friel tell Paul Weller they are delighted with his version of El Dorado.15

Russia is vowing Thursday to repair the warplanes damaged by Ukraine in a massive drone attack earlier this week, with an official claiming they were "not destroyed but damaged." The comments from Russian Deputy Foreign Minister Sergey Ryabkov come after Ukraine said its forces destroyed 40 of Russia's most powerful bomber jets and surveillance planes in "Operation Spider's Web," a series of coordinated drone strikes Sunday penetrating deep into Russian territory. "As the defense ministry said, these aircraft were not destroyed but damaged. They will be repaired," Ryabkov was quoted telling Russia's state-run TASS news agency. However, satellite images of Russian airfields show extensive damage to the planes.