Transformers, especially the decoder-only variants, are the backbone of most modern large language models; yet we do not have much understanding of their expressive power except for the simple $1$-layer case. Due to the difficulty of analyzing multi-layer models, all previous work relies on unproven complexity conjectures to show limitations for multi-layer Transformers. In this work, we prove the first $\textit{unconditional}$ lower bound against multi-layer decoder-only transformers. For any constant $L$, we prove that any $L$-layer decoder-only transformer needs a polynomial model dimension ($n^{\Omega(1)}$) to perform sequential composition of $L$ functions over an input of $n$ tokens. As a consequence, our results give: (1) the first depth-width trade-off for multi-layer transformers, exhibiting that the $L$-step composition task is exponentially harder for $L$-layer models compared to $(L+1)$-layer ones; (2) an unconditional separation between encoder and decoder, exhibiting a hard task for decoders that can be solved by an exponentially shallower and smaller encoder; (3) a provable advantage of chain-of-thought, exhibiting a task that becomes exponentially easier with chain-of-thought. On the technical side, we propose the multi-party $\textit{autoregressive}$ $\textit{communication}$ $\textit{model}$ that captures the computation of a decoder-only Transformer. We also introduce a new proof technique that finds a certain $\textit{indistinguishable}$ $\textit{decomposition}$ of all possible inputs iteratively for proving lower bounds in this model. We believe our new communication model and proof technique will be helpful to further understand the computational power of transformers.

Recently, multimodal large language models (MLLMs), such as GPT-4o, Gemini 1.5 Pro, and Reka Core, have expanded their capabilities to include vision and audio modalities. While these models demonstrate impressive performance across a wide range of audio-visual applications, our proposed DeafTest reveals that MLLMs often struggle with simple tasks humans find trivial: 1) determining which of two sounds is louder, and 2) determining which of two sounds has a higher pitch. Motivated by these observations, we introduce AV-Odyssey Bench, a comprehensive audio-visual benchmark designed to assess whether those MLLMs can truly understand the audio-visual information. This benchmark encompasses 4,555 carefully crafted problems, each incorporating text, visual, and audio components. To successfully infer answers, models must effectively leverage clues from both visual and audio inputs. To ensure precise and objective evaluation of MLLM responses, we have structured the questions as multiple-choice, eliminating the need for human evaluation or LLM-assisted assessment. We benchmark a series of closed-source and open-source models and summarize the observations. By revealing the limitations of current models, we aim to provide useful insight for future dataset collection and model development.

Narrative understanding and story generation are critical challenges in natural language processing (NLP), with much of the existing research focused on summarization and question-answering tasks. While previous studies have explored predicting plot endings and generating extended narratives, they often neglect the logical coherence within stories, leaving a significant gap in the field. To address this, we introduce the Missing Logic Detector by Emotion and Action (MLD-EA) model, which leverages large language models (LLMs) to identify narrative gaps and generate coherent sentences that integrate seamlessly with the story's emotional and logical flow. The experimental results demonstrate that the MLD-EA model enhances narrative understanding and story generation, highlighting LLMs' potential as effective logic checkers in story writing with logical coherence and emotional consistency. This work fills a gap in NLP research and advances border goals of creating more sophisticated and reliable story-generation systems.

Recent advancements in large vision-language models (LVLM) have significantly enhanced their ability to comprehend visual inputs alongside natural language. However, a major challenge in their real-world application is hallucination, where LVLMs generate non-existent visual elements, eroding user trust. The underlying mechanism driving this multimodal hallucination is poorly understood. Minimal research has illuminated whether contexts such as sky, tree, or grass field involve the LVLM in hallucinating a frisbee. We hypothesize that hidden factors, such as objects, contexts, and semantic foreground-background structures, induce hallucination. This study proposes a novel causal approach: a hallucination probing system to identify these hidden factors. By analyzing the causality between images, text prompts, and network saliency, we systematically explore interventions to block these factors. Our experimental findings show that a straightforward technique based on our analysis can significantly reduce hallucinations. Additionally, our analyses indicate the potential to edit network internals to minimize hallucinated outputs.

This paper introduces the human-curated PandasPlotBench dataset, designed to evaluate language models' effectiveness as assistants in visual data exploration. Our benchmark focuses on generating code for visualizing tabular data - such as a Pandas DataFrame - based on natural language instructions, complementing current evaluation tools and expanding their scope. The dataset includes 175 unique tasks. Our experiments assess several leading Large Language Models (LLMs) across three visualization libraries: Matplotlib, Seaborn, and Plotly. We show that the shortening of tasks has a minimal effect on plotting capabilities, allowing for the user interface that accommodates concise user input without sacrificing functionality or accuracy. Another of our findings reveals that while LLMs perform well with popular libraries like Matplotlib and Seaborn, challenges persist with Plotly, highlighting areas for improvement. We hope that the modular design of our benchmark will broaden the current studies on generating visualizations. Our benchmark is available online: https://huggingface.co/datasets/JetBrains-Research/plot_bench. The code for running the benchmark is also available: https://github.com/JetBrains-Research/PandasPlotBench.

Low-Rank Adaptation (LoRA) and other parameter-efficient fine-tuning (PEFT) methods provide low-memory, storage-efficient solutions for personalizing text-to-image models. However, these methods offer little to no improvement in wall-clock training time or the number of steps needed for convergence compared to full model fine-tuning. While PEFT methods assume that shifts in generated distributions (from base to fine-tuned models) can be effectively modeled through weight changes in a low-rank subspace, they fail to leverage knowledge of common use cases, which typically focus on capturing specific styles or identities. Observing that desired outputs often comprise only a small subset of the possible domain covered by LoRA training, we propose reducing the search space by incorporating a prior over regions of interest. We demonstrate that training a hypernetwork model to generate LoRA weights can achieve competitive quality for specific domains while enabling near-instantaneous conditioning on user input, in contrast to traditional training methods that require thousands of steps.

Large language models (LLMs) have shown remarkable potential as autonomous agents, particularly in web-based tasks. However, existing LLM web agents heavily rely on expensive proprietary LLM APIs, while open LLMs lack the necessary decision-making capabilities. This paper introduces WebRL, a self-evolving online curriculum reinforcement learning framework designed to train high-performance web agents using open LLMs. WebRL addresses three key challenges in building LLM web agents, including the scarcity of training tasks, sparse feedback signals, and policy distribution drift in online learning. Specifically, WebRL incorporates 1) a self-evolving curriculum that generates new tasks from unsuccessful attempts, 2) a robust outcome-supervised reward model (ORM), and 3) adaptive reinforcement learning strategies to ensure consistent improvements. We apply WebRL to transform open Llama-3.1 and GLM-4 models into proficient web agents. On WebArena-Lite, WebRL improves the success rate of Llama-3.1-8B from 4.8% to 42.4%, and from 6.1% to 43% for GLM-4-9B. These open models significantly surpass the performance of GPT-4-Turbo (17.6%) and GPT-4o (13.9%) and outperform previous state-of-the-art web agents trained on open LLMs (AutoWebGLM, 18.2%). Our findings demonstrate WebRL's effectiveness in bridging the gap between open and proprietary LLM-based web agents, paving the way for more accessible and powerful autonomous web interaction systems.

Over the past decade, the revolution in single-cell sequencing has enabled the simultaneous molecular profiling of various modalities across thousands of individual cells, allowing scientists to investigate the diverse functions of complex tissues and uncover underlying disease mechanisms. Among all the analytical steps, assigning individual cells to specific types is fundamental for understanding cellular heterogeneity. However, this process is usually labor-intensive and requires extensive expert knowledge. Recent advances in large language models (LLMs) have demonstrated their ability to efficiently process and synthesize vast corpora of text to automatically extract essential biological knowledge, such as marker genes, potentially promoting more efficient and automated cell type annotations. To thoroughly evaluate the capability of modern instruction-tuned LLMs in automating the cell type identification process, we introduce SOAR, a comprehensive benchmarking study of LLMs for cell type annotation tasks in single-cell genomics. Specifically, we assess the performance of 8 instruction-tuned LLMs across 11 datasets, spanning multiple cell types and species. Our study explores the potential of LLMs to accurately classify and annotate cell types in single-cell RNA sequencing (scRNA-seq) data, while extending their application to multiomics data through cross-modality translation. Additionally, we evaluate the effectiveness of chain-of-thought (CoT) prompting techniques in generating detailed biological insights during the annotation process. The results demonstrate that LLMs can provide robust interpretations of single-cell data without requiring additional fine-tuning, advancing the automation of cell type annotation in genomics research.

To address this, conditioning methods have been proposed, such as ControlNet [51] and IPadapter [48], that aim to capture the desired shape or form more explicitly through the use of edge or depth maps as input conditions. Despite these advancements, current text-and image-conditioned synthesis approaches still face a number of challenges. First, they often struggle to balance both textual and visual conditions, when text describes a particular context that should be combined with the target shape to guide an image synthesis (Figure 1, top row). Second, commonly used visual conditions such as edge or depth maps are limited to a single viewpoint, resulting in a loss of valuable 3D shape information when users seek image variations of an underlying shape from different poses. Third, even when these models accurately reflect the target shape in specific views, users may want to explore shape variations - yet current models often lack flexible controls for such exploration. To overcome these challenges, we propose ShapeWords, a method designed to generate images that faithfully adhere to both the text prompt and a target 3D shape geometry, 1 arXiv:2412.02912v1

We introduce FathomGPT, an open source system for the interactive investigation of ocean science data via a natural language interface. FathomGPT was developed in close collaboration with marine scientists to enable researchers to explore and analyze the FathomNet image database. FathomGPT provides a custom information retrieval pipeline that leverages OpenAI's large language models to enable: the creation of complex queries to retrieve images, taxonomic information, and scientific measurements; mapping common names and morphological features to scientific names; generating interactive charts on demand; and searching by image or specified patterns within an image. In designing FathomGPT, particular emphasis was placed on enhancing the user's experience by facilitating free-form exploration and optimizing response times. We present an architectural overview and implementation details of FathomGPT, along with a series of ablation studies that demonstrate the effectiveness of our approach to name resolution, fine tuning, and prompt modification. We also present usage scenarios of interactive data exploration sessions and document feedback from ocean scientists and machine learning experts.