Research→Official→arXiv Machine Learning
This study systematically compares Turbo-Quant and SpectralQuant KV-cache compression methods, using a statistical validation methodology to evaluate non-dominated schemes. The research finds that eigenbasis-based methods perform poorly on heavy-tailed data due to covariance instability but excel in structured regimes, with the effective semantic dimension adapting to calibration budgets rather than true data rank.
Why it matters: The work provides a rigorous framework for evaluating KV-cache compression techniques, which is important for optimizing memory and latency in large language model inference.
Jul 14, 2026
Research→Official→arXiv Information Retrieval
Researchers introduce PaperRouter-Agent, a training-free large language model (LLM) agent designed to route new academic papers into user-specific folder hierarchies by grounding its decisions in the content of folder members rather than relying solely on folder names. In tests on real personal libraries, PaperRouter-Agent improved Recall@1 from 0.39 to 0.61 and Recall@3 from 0.57 to 0.83. On the public LaMP-2 benchmark, it increased accuracy from 44.5% to 51.5% and macro-F1 by 9.0 points over a single-shot baseline.
Why it matters: This work demonstrates a significant advance in personalized information organization, showing that LLM agents can effectively route papers in reference managers without per-user training.
Jul 14, 2026
Research→Official→arXiv Information Retrieval
Researchers introduce TALRanker, a framework that models relevance scoring as a Markov decision process, enabling large language models to selectively use external tools only when uncertain. The approach employs a two-stage training process—first preventing catastrophic forgetting, then using reinforcement learning to optimize tool invocation. TALRanker achieves state-of-the-art results on retrieval benchmarks while maintaining throughput comparable to pointwise rerankers.
Why it matters: This work offers a novel solution to the accuracy-efficiency trade-off in LLM-based reranking by enabling models to autonomously decide when to use external tools, reducing latency without compromising performance.
Jul 14, 2026
Research→Official→arXiv Machine Learning
A new preprint introduces Low-Rank Attention Residuals (LR-AttnRes), a method that uses low-dimensional keys for depthwise routing in large language models (LLMs) while retaining full-dimensional residual values. This approach decouples routing from representation, leading to improved validation loss and reduced computational cost. The authors present two variants—Projected and Sliced LR-AttnRes—and release code and models for further research.
Why it matters: The work demonstrates that effective depthwise routing in LLMs can be achieved with significantly fewer dimensions, suggesting a path toward more efficient model architectures.
Jul 14, 2026
Models→Official→arXiv Information Retrieval
Meta introduces RankGraph-2, a framework that co-designs graph construction, representation learning, and real-time serving for similarity-based retrieval at billion-node scale. The system reduces serving computational cost by 83% and achieves up to +0.96% click-through rate (CTR) and +2.75% conversion rate (CVR) improvements. RankGraph-2 has been deployed to power over 20 retrieval launches across major Meta surfaces.
Why it matters: RankGraph-2 demonstrates a significant advance in large-scale recommendation systems by jointly optimizing all stages of graph-based retrieval, resulting in notable efficiency and performance gains.
Jul 14, 2026
Research→Official→arXiv Information Retrieval
Researchers have introduced Adaptive Model Compression (AMC), a framework that dynamically allocates hardware resources during transformer inference based on token saliency. AMC uses a multi-tier architecture to process important tokens at full precision while compressing less significant data, resulting in a 59.2% reduction in system energy and a 2.24x increase in throughput on 45nm CMOS hardware, with only a 3.6% drop in accuracy.
Why it matters: This approach could make it feasible to run large transformer models efficiently on edge devices, significantly improving battery life without major performance loss.
Jul 14, 2026
Research→Official→arXiv Machine Learning
EvoClawBench is a new benchmark designed to test whether AI agents can transform evidence from their own runs into reusable skills that improve future executions. The benchmark covers 100 tasks across coding, data, office, security, operations, and domain-document workflows, and supports multiple agent runtimes. Experiments with OpenClaw and nanobot show that the ability to learn reusable skills is selective and cost-sensitive: some models improve with skill learning, while others experience performance drops or collapse. This demonstrates that skill learning from experience is not an automatic benefit for agent systems.
Why it matters: This work isolates and rigorously tests a critical capability for autonomous agents—learning from their own experience—and shows that it is not guaranteed, informing the design of more robust agent systems.
Jul 14, 2026
Research→Official→arXiv Machine Learning
A new preprint introduces SPARC, a spectral-algebraic theory that formally explains why autoregressive language models (LLMs) often fail to correct their own errors but can fix identical errors when attributed to external sources. The theory proves that this 'self-correction blind spot' arises when the spectral radius of the error-propagation operator is at least one, and it derives a precise activation threshold for correction markers. Experiments across four model backbones and a visual autoregressive probe validate the theory, with predictions matching observed blind-spot rates within 3.2% RMSE.
Why it matters: This work provides the first formal, quantitative model of the self-correction blind spot in autoregressive generation, offering new insights that could improve the reliability of LLMs and related models.
Jul 14, 2026
Policy Safety→Official→arXiv Computers and Society
A new preprint argues that as AI models approach top performance on existing benchmarks, the remaining discriminating items are those requiring elite expert judgment, which is structurally scarce. The authors present a formal model showing how benchmark signal depreciates as models improve, and document a scarcity premium for high-judgment evaluation labor. The paper also discusses the governance implications of these findings for AI capability measurement.
Why it matters: This work highlights a fundamental bottleneck in evaluating advanced AI systems, raising concerns about the reliability of current benchmarks and the challenges for AI governance.
Jul 14, 2026
Research→Official→arXiv Information Retrieval
Researchers present CORE-Bench, a new benchmark designed to evaluate code retrieval in agentic coding scenarios. The benchmark assesses models on code understanding, issue-to-edit localization, and broader context retrieval, using over 180,000 queries and 106,000 relevance labels. Results show that current embedding models perform poorly on these tasks, but simple supervised fine-tuning leads to notable improvements, highlighting significant room for advancement.
Why it matters: CORE-Bench fills a critical gap in evaluating code retrieval for AI coding agents, supporting the development of more capable systems for navigating and understanding code repositories.
Jul 14, 2026
Research→Official→arXiv Information Retrieval
Researchers introduce GRASP, a reinforcement learning framework that enables agents to adaptively coordinate semantic search, keyword search, and paragraph reading during multi-step reasoning. The policy learns to control context granularity, leading to improved retrieval recall and question answering performance on multi-hop benchmarks. The learned strategies include interpretable skimming and scanning behaviors.
Why it matters: This work advances agentic retrieval-augmented generation by enabling dynamic and context-aware retrieval strategies, which are important for accurate multi-step reasoning.
Jul 14, 2026
Research→Official→arXiv Machine Learning
A new preprint evaluates Scientific Machine Learning (SciML) methods for macroeconomic forecasting and finds that structural priors can act as misregularizers when they do not align with the true data-generating process. In tests across 23 countries, less-constrained models like ARIMA and Neural ODEs consistently outperformed more-constrained models such as PINNs and UDEs. The study highlights the challenges of low-frequency macroeconomic prediction and cautions that more structural assumptions do not always yield better results.
Why it matters: This work challenges the assumption that adding structural priors always improves model performance, offering important guidance for SciML practitioners.
Jul 14, 2026
Research→Official→arXiv Information Retrieval
SVD-RAG introduces the use of Singular Value Decomposition (SVD) on dense sentence embeddings for extractive summarization in hierarchical Retrieval-Augmented Generation (RAG) systems, replacing the need for expensive LLM-based summarization. The method achieves retrieval quality within 1-5% of RAPTOR while constructing the retrieval tree 317 times faster and reducing token consumption by approximately 85%. SVD-RAG is deterministic, cost-efficient, and adapts to content complexity automatically.
Why it matters: This approach makes hierarchical RAG systems significantly more practical and scalable by reducing computational cost and latency without substantially sacrificing retrieval quality.
Jul 14, 2026
Research→Official→arXiv Information Retrieval
FAIR GraphRAG is a novel framework that incorporates FAIR Digital Objects as nodes in a graph-based retrieval system to enhance retrieval-augmented generation for domain-specific question answering. Co-designed by physicians and computer scientists, the system leverages large language models for schema construction and metadata extraction, and demonstrates improved accuracy, coverage, and explainability on complex biomedical queries involving metadata and ontology links. The approach is validated on a gastroenterology RNA-sequencing dataset and is positioned as applicable to other specialized domains.
Why it matters: This work represents a meaningful advance by combining FAIR data principles with graph-based retrieval and LLMs, improving semantic data analysis and question answering in complex, specialized fields.
Jul 14, 2026
Research→Official→arXiv Computers and Society
A new preprint explores 'model collapse,' the phenomenon where recursively training AI models on AI-generated data leads to degraded performance, such as repetition and noise. The author argues that while this is typically seen as a technical failure, it also has creative and aesthetic dimensions, drawing parallels to early video feedback art. The paper examines how model collapse challenges certain technological ideals and underscores AI's ongoing reliance on human-generated data.
Why it matters: This research reframes model collapse as not only a technical issue but also a source of artistic and philosophical insight, broadening our understanding of AI's creative potential and limitations.
Jul 14, 2026
Research→Official→arXiv Information Retrieval
Researchers introduce Prompt Generation (PG), a configuration-driven framework that decouples feature-processing logic from model architecture using declarative JSON files. This approach standardizes feature processing, enabling faster training iteration, streamlined deployment, and efficient online inference. Deployed on Taobao Search, PG achieved statistically significant uplifts of +0.47% in transaction count and +0.51% in gross merchandise value (GMV) during online A/B tests.
Why it matters: PG demonstrates a practical advance in industrial AI systems by reducing engineering complexity and accelerating deployment for large-scale search and recommendation platforms.
Jul 14, 2026
Policy Safety→Official→arXiv Cryptography and Security
A new preprint demonstrates that adversaries can corrupt open datasets to poison AI-driven scientific research, with attacks succeeding in nearly half of experimental runs and detection rates remaining very low. The attack leverages autonomous research agents that retrieve and process public data, potentially turning well-intentioned scientists into unwitting distributors of fraudulent results. The study also finds that implementing provenance auditing with five specific checks can reduce attack success to zero.
Why it matters: This research exposes a scalable vulnerability in AI-assisted science that could industrialize scientific fraud, posing a significant threat to the integrity of research in critical domains.
Jul 14, 2026
Policy Safety→Official→arXiv Cryptography and Security
A new preprint demonstrates that simple prompt reframing—such as presenting questions as medical board exam items—can bypass safety guardrails in Google's MedGemma-4B medical language model. The study found an overall attack success rate of 38.0%, with the drug-interaction guardrail being particularly weak (83.2% ASR) and the emergency-deferral guardrail remaining robust (4.7% ASR).
Why it matters: This work reveals that current safety guardrails in open-weight medical language models can be circumvented by unsophisticated prompt engineering, raising concerns about their reliability in real-world applications.
Jul 14, 2026
Research→Official→arXiv Information Retrieval
A new preprint introduces MC-RAG, a structure-driven retrieval-augmented generation (RAG) system that reformulates retrieval as a subgraph matching problem over a knowledge graph. By combining semantic and structural embeddings with path-level indexing, MC-RAG aims to improve the handling of complex, multi-constraint queries, offering more interpretable and constraint-consistent retrieval and generation. The system is demonstrated with interactive examples and a demo video.
Why it matters: MC-RAG proposes a novel approach to address the challenge of constraint violations and hallucinations in RAG systems when handling complex queries, potentially improving reliability and interpretability.
Jul 14, 2026
Research→Official→arXiv Information Retrieval
A new preprint demonstrates that cross-lingual prompt injection attacks can significantly inflate relevance scores in LLM-based information retrieval systems, while evading current prompt-injection defenses. The attacks, tested across eight languages and multiple open-weight models, can also adapt to bypass modified defense mechanisms. This exposes a critical vulnerability in using LLMs as automated relevance judges.
Why it matters: The study reveals that language diversity can be exploited as an attack vector, exposing a major security gap in LLM-based evaluation systems.
Jul 14, 2026