Large language models (LLMs) are increasingly deployed with task-specific adapters catering to multiple downstream applications. In such a scenario, the additional compute associated with these apparently insignificant number of adapter parameters (typically less than 1% of the base model) turns out to be disproportionately significant during inference time (upto 2.5x times that of the base model). In this paper, we propose a new zero-latency fused low-rank adapter (zFLoRA) that introduces zero or negligible latency overhead on top of the base model. Experimental results on LLMs of size 1B, 3B and 7B show that zFLoRA compares favorably against the popular supervised fine-tuning benchmarks including low-rank adapters (LoRA) as well as full fine-tuning (FFT). Experiments are conducted on 18 different tasks across three different categories namely commonsense reasoning, math reasoning and summary-dialogue. Latency measurements made on NPU (Samsung Galaxy S25+) as well as GPU (NVIDIA H100) platforms show that the proposed zFLoRA adapters introduce zero to negligible latency overhead.

As LLM agents advance, they are increasingly mediating economic decisions, ranging from product discovery to transactions, on behalf of users. Such applications promise benefits but also raise many questions about agent accountability and value for users. Addressing these questions requires understanding how agents behave in realistic market conditions. However, previous research has largely evaluated agents in constrained settings, such as single-task marketplaces (e.g., negotiation) or structured two-agent interactions. Real-world markets are fundamentally different: they require agents to handle diverse economic activities and coordinate within large, dynamic ecosystems where multiple agents with opaque behaviors may engage in open-ended dialogues. To bridge this gap, we investigate two-sided agentic marketplaces where Assistant agents represent consumers and Service agents represent competing businesses. To study these interactions safely, we develop Magentic-Marketplace -- a simulated environment where Assistants and Services can operate. This environment enables us to study key market dynamics: the utility agents achieve, behavioral biases, vulnerability to manipulation, and how search mechanisms shape market outcomes. Our experiments show that frontier models can approach optimal welfare -- but only under ideal search conditions. Performance degrades sharply with scale, and all models exhibit severe first-proposal bias, creating 10-30x advantages for response speed over quality. These findings reveal how behaviors emerge across market conditions, informing the design of fair and efficient agentic marketplaces.

There is a substantial body of literature examining the mathematical reasoning capabilities of large language models (LLMs), particularly their performance on precise arithmetic operations in autoregressive architectures. However, their ability to perform approximate reasoning in informal, fast-paced mathematical operations has received far less attention, especially among non-autoregressive decoder models. Our work addresses this gap by introducing StreetMath, a benchmark designed to evaluate models' approximation abilities under real-world approximation scenarios. We conduct extensive evaluations across different LLM architectures: Qwen3-4B-Instruct-2507, Qwen3-4B-Thinking-2507, Dream-v0-Instruct-7B, Falcon-Mamba-7B-Instruct, and Mamba-GPT-3B. Furthermore, we apply mechanistic interpretability techniques to probe their internal computational states. Our analysis reveals that LLMs generally attempt to compute exact values or invoke external tools even in tasks that call for approximation. Moreover, while models sometimes reach the correct answer in early layers or steps, they still consume more tokens when solving approximation tasks. Additional experiments indicate that exact and approximate arithmetic operations rely on largely separate neural components. Drawing upon research on cognitive psychology, we argue that LLMs do not exhibit cognitive miserliness in the same way humans do in street math settings. We open source our work https://github.com/ctseng777/StreetMath

Current evaluations of agents remain centered around one-shot task completion, failing to account for the inherently iterative and collaborative nature of many real-world problems, where human goals are often underspecified and evolve. We argue for a shift from building and assessing task completion agents to developing collaborative agents, assessed not only by the quality of their final outputs but by how well they engage with and enhance human effort throughout the problem-solving process. To support this shift, we introduce collaborative effort scaling, a framework that captures how an agent's utility grows with increasing user involvement. Through case studies and simulated evaluations, we show that state-of-the-art agents often underperform in multi-turn, real-world scenarios, revealing a missing ingredient in agent design: the ability to sustain engagement and scaffold user understanding. Collaborative effort scaling offers a lens for diagnosing agent behavior and guiding development toward more effective interactions.

Unified vision-language models (UVLMs) must perform both understanding and generation within a single architecture, but these tasks rely on heterogeneous data and supervision, making it difficult to balance them during reinforcement learning (RL). We propose PairUni, a unified framework that reorganizes data into understanding-generation (UG) pairs and aligns optimization accordingly. We first use GPT-o3 to augment single-task data, generating captions for understanding samples and question-answer (QA) pairs for generation samples, forming aligned pairs from the same instance. Additionally, for each generation sample, we retrieve a semantically related understanding example to form a retrieved pair, linking different but related data points. These paired structures expose cross-task semantic correspondences and support consistent policy learning. To leverage this structure, we present Pair-GPRO, a pair-aware variant based on Group Relative Policy Optimization. It assigns a similarity score to each pair to modulate the advantage, strengthening learning from well-aligned examples and reducing task interference. We curate a high-quality dataset of 16K UG pairs named PairUG for RL fine-tuning and evaluate PairUni on the powerful Janus-Pro UVLMs. Our approach achieves balanced improvements on various UVLMs, outperforming strong UVLM RL baselines. Codes are available at https://github.com/Haochen-Wang409/PairUni.

Test-time scaling (TTS) techniques can improve the performance of large language models (LLMs) at the expense of additional computation and latency. While TTS has proven effective in formal domains such as mathematics and programming, its value in argumentative domains such as law remains underexplored. We present an empirical study of verifier-based TTS methods for legal multiple-choice QA (MCQA) across five benchmarks. Using a family of 7 reward models, we evaluate both outcome-level (Best-of-$N$) and process-level (tree search) verification under realistic low-$N$ budgets. Our analysis systematically investigates how verifier utility is affected by key properties such as domain specialization, model size, and supervision type (process-supervised PRMs vs. outcome-only ORMs), even when applied across different roles.

Large Language Models (LLMs) are exhibiting emergent human-like abilities and are increasingly envisioned as the foundation for simulating an individual's communication style, behavioral tendencies, and personality traits. However, current evaluations of LLM-based persona simulation remain limited: most rely on synthetic dialogues, lack systematic frameworks, and lack analysis of the capability requirement. To address these limitations, we introduce TwinVoice, a comprehensive benchmark for assessing persona simulation across diverse real-world contexts. TwinVoice encompasses three dimensions: Social Persona (public social interactions), Interpersonal Persona (private dialogues), and Narrative Persona (role-based expression). It further decomposes the evaluation of LLM performance into six fundamental capabilities, including opinion consistency, memory recall, logical reasoning, lexical fidelity, persona tone, and syntactic style. Experimental results reveal that while advanced models achieve moderate accuracy in persona simulation, they still fall short of capabilities such as syntactic style and memory recall. Consequently, the average performance achieved by LLMs remains considerably below the human baseline.

The rapid advancement of large language models(LLMs) has intensified the need for domain and culture specific evaluation. Existing benchmarks are largely Anglocentric and domain-agnostic, limiting their applicability to India-centric contexts. To address this gap, we introduce BhashaBench V1, the first domain-specific, multi-task, bilingual benchmark focusing on critical Indic knowledge systems. BhashaBench V1 contains 74,166 meticulously curated question-answer pairs, with 52,494 in English and 21,672 in Hindi, sourced from authentic government and domain-specific exams. It spans four major domains: Agriculture, Legal, Finance, and Ayurveda, comprising 90+ subdomains and covering 500+ topics, enabling fine-grained evaluation. Evaluation of 29+ LLMs reveals significant domain and language specific performance gaps, with especially large disparities in low-resource domains. For instance, GPT-4o achieves 76.49% overall accuracy in Legal but only 59.74% in Ayurveda. Models consistently perform better on English content compared to Hindi across all domains. Subdomain-level analysis shows that areas such as Cyber Law, International Finance perform relatively well, while Panchakarma, Seed Science, and Human Rights remain notably weak. BhashaBench V1 provides a comprehensive dataset for evaluating large language models across India's diverse knowledge domains. It enables assessment of models' ability to integrate domain-specific knowledge with bilingual understanding. All code, benchmarks, and resources are publicly available to support open research.

Recent advances in text-to-video (T2V) generation highlight the critical role of high-quality video-text pairs in training models capable of producing coherent and instruction-aligned videos. However, strategies for optimizing video captions specifically for T2V training remain underexplored. In this paper, we introduce VC4VG (Video Captioning for Video Generation), a comprehensive caption optimization framework tailored to the needs of T2V models. We begin by analyzing caption content from a T2V perspective, decomposing the essential elements required for video reconstruction into multiple dimensions, and proposing a principled caption design methodology. To support evaluation, we construct VC4VG-Bench, a new benchmark featuring fine-grained, multi-dimensional, and necessity-graded metrics aligned with T2V-specific requirements. Extensive T2V fine-tuning experiments demonstrate a strong correlation between improved caption quality and video generation performance, validating the effectiveness of our approach. We release all benchmark tools and code at https://github.com/alimama-creative/VC4VG to support further research.

The task of information extraction (IE) is to extract structured knowledge from text. However, it is often not straightforward to utilize IE output due to the mismatch between the IE ontology and the downstream application needs. We propose a new formulation of IE TEXT2DB that emphasizes the integration of IE output and the target database (or knowledge base). Given a user instruction, a document set, and a database, our task requires the model to update the database with values from the document set to satisfy the user instruction. This task requires understanding user instructions for what to extract and adapting to the given DB/KB schema for how to extract on the fly. To evaluate this new task, we introduce a new benchmark featuring common demands such as data infilling, row population, and column addition. In addition, we propose an LLM agent framework OPAL (Observe-PlanAnalyze LLM) which includes an Observer component that interacts with the database, the Planner component that generates a code-based plan with calls to IE models, and the Analyzer component that provides feedback regarding code quality before execution. Experiments show that OPAL can successfully adapt to diverse database schemas by generating different code plans and calling the required IE models. We also highlight difficult cases such as dealing with large databases with complex dependencies and extraction hallucination, which we believe deserve further investigation. Source code: https://github.com/yzjiao/Text2DB