Large language models (LLMs) have demonstrated rapidly improving long-context capabilities, prompting a wave of benchmarks designed to evaluate them. However, existing long-context evaluations - from Needle-in-a-Haystack (NIAH) tests to more recent multi-hop reasoning and summarization tasks - predominantly measure average-case performance, and many are either saturated or lack robustness.
While Large Language Models (LLMs) have become essential productivity tools, their integration into workflows without adequate safeguards creates significant risks. This paper proposes an open-source, privacy-focused, user-facing firewall designed to secure both web-based and programmatic LLM interactions.
Large language model (LLM)-based lossless image compression methods typically represent pixel data through the native text interface of a pretrained model, converting pixel values into token sequences that the LLM processes through its vocabulary head. This design shows that pretrained language models can provide probability estimates for image coding, but it also couples compression to tokenizer behavior, vocabulary-specific numeric tokens, and model-family-specific adaptation.
Scaling Large Language Models (LLMs) has been driven mainly by enlarging the Transformer backbone, but for an already-strong model this requires another round of costly pretraining. We study whether an existing backbone can keep improving by allocating more computation to each token while leaving the Transformer backbone fixed.
Deep neural nets achieve remarkable performance when training and test data share the same distribution, but this assumption frequently breaks in real-world deployment, where data undergoes continual distributional shifts. Continual Test-Time Adaptation (CTTA) addresses this challenge by adapting pretrained models to non-stationary target distributions on-the-fly, without access to source data or labeled targets, while mitigating two critical failure modes: catastrophic forgetting of source knowledge and error accumulation from noisy pseudo-labels over extended time horizons.
Large language models (LLMs) increasingly act as integrated data-science agents, combining abstract reasoning with advanced tool use. Yet the relevant benchmark landscape largely divides into symbolic causal reasoning benchmarks without realistic data analysis or data analysis benchmarks without a principled causal data-generating structure.
arXiv:2605. 15207v2 Announce Type: replace Abstract: Multi-agent LLM systems have shown promise for complex reasoning, yet recent evaluations reveal they often underperform single-model baselines.
By Yi Xie, Siao Liu, Falong Fan, Yuanqi Yao, Yue Zhao, Bo Liu
arXiv:2604. 00533v2 Announce Type: replace Abstract: Large Language Models (LLMs) generalize across tasks through reusable representations and flexible reasoning, yet remain brittle in real deployment when faced with evolving tasks and continual distribution shift.
By Xiao Zhang, Tianyu Hu, Juntao Lyu, Qianchuan Zhao, Huimin Ma
arXiv:2607. 07386v1 Announce Type: new Abstract: Linear attention models allow a fixed state size and a fixed amount of compute per token.
By Lo\"ic Cabannes, Pierre-Emmanuel Mazar\'e, Gergely Szilvasy, Matthijs Douze, Maria Lomeli, Ilze Amanda Auzina, Justin Carpentier, Gabriel Synnaeve, Herv\'e J\'egou
arXiv:2607. 07026v1 Announce Type: new Abstract: Constrained decoding is essential for serving LLMs, ensuring that generated outputs follow specific structures such as JSON schema-formatted function calls.
By Meihua Dang, Stefano Ermon
arXiv:2510. 04547v5 Announce Type: replace Abstract: Large pretrained vision encoders are central to multimodal intelligence, powering applications from on-device vision processing to vision-language models.
By Seunghyeon Kim, Taesun Yeom, Jinho Kim, Wonpyo Park, Kyuyeun Kim, Jaeho Lee
arXiv:2607. 06987v1 Announce Type: new Abstract: Reinforcement learning (RL) has become the standard paradigm for enhancing the complex reasoning capabilities of large language models (LLMs).
By Chongyu Fan, Pengfei Liu, Jingjia Huang, Sijia Liu, Yi Lin
arXiv:2607. 07375v1 Announce Type: cross Abstract: Adversarial vulnerability in deep neural networks (DNNs) has been studied from the perspectives of decision-boundary geometry, feature robustness, input-output Jacobians, and the instability of inverse problems.
By Chethan Krishnamurthy Ramanaik, Tobias Callies, Michael Hecht, Eirini Ntoutsi
arXiv:2607. 07626v1 Announce Type: cross Abstract: Reliable confidence estimation is essential for deploying large language models (LLMs) in confidence-aware systems, where downstream decisions such as retrieval, tool use, and adaptive computation depend on accurately estimating answer reliability.
By Sahil Kale
arXiv:2607. 06653v1 Announce Type: new Abstract: Federated learning (FL) enables collaborative model training across institutions without sharing sensitive patient data.
By Kien Le, Joseph Lindley, Quoc Bao Phan, Tuy Tan Nguyen
arXiv:2607. 07646v1 Announce Type: new Abstract: Does RL post-training merely amplify primitive skills already latent in a base model, or can it compose primitive skills into new higher-level strategies?
By Azwar Abdulsalam, Nishil Patel, Andrew Saxe
arXiv:2607. 06620v1 Announce Type: cross Abstract: Recent Multimodal Large Language Models (MLLMs) struggle to bridge the representational gap between 2D semantic understanding and 3D spatial geometry.
By Haida Feng, Hao Wei, Haolin Wang, Shiwei Li, Chade Li, Yihong Wu
arXiv:2607. 07696v1 Announce Type: cross Abstract: Analytical workloads operating on data stored in external database systems face a fundamental bottleneck: data access is guarded entirely by the database driver, like JDBC or ODBC, forcing all reads through query execution and other driver layers that are not designed for bulk columnar analytics.
By Victor Giannakouris, Immanuel Trummer
arXiv:2607. 07117v1 Announce Type: cross Abstract: In text-to-image in-context learning (T2I-ICL), a model has to infer a latent compositional pattern from fewshot demonstrations for generating a query image.
By Stepanida Alekseeva, Jenifer Kalafatovich, Seong-Whan Lee
arXiv:2605. 11325v3 Announce Type: replace-cross Abstract: Current LLM memory benchmarks evaluate answer quality rather than retrieval accuracy.
By Jeffrey Flynt