Adaptive Memory Decay for Log-Linear Attention
arXiv:2605.06946v1 Announce Type: cross Abstract: Sequence models face a fundamental tradeoff between memory capacity and computational efficiency. Transformers achieve expressive context modeling at
Knowledge catalogue
arXiv:2605.06946v1 Announce Type: cross Abstract: Sequence models face a fundamental tradeoff between memory capacity and computational efficiency. Transformers achieve expressive context modeling at
arXiv:2602.22831v2 Announce Type: replace-cross Abstract: Moral benchmarks for LLMs typically score models on context-free prompts, implicitly treating the measured choice rate as stable. We test this
arXiv:2605.07565v1 Announce Type: cross Abstract: We study zeroth-order optimisation under context distributional uncertainty, a setting commonly tackled using Bayesian optimisation (BO). A prevailing
arXiv:2605.07357v1 Announce Type: new Abstract: Reasoning-acting frameworks enhance large language models (LLMs) by interleaving reasoning with actions for dynamic information acquisition. However, ex
arXiv:2602.06283v2 Announce Type: replace Abstract: Exploiting sparsity during long-context inference is key to scaling large language models, as attention dominates the cost of autoregressive decodin
Swyx posted a tweet referencing Vivian Bala with a statement about demonstrating something to the world, though the specific context of what will be demonstrated is not clear from the title alone. The
DeepSeek released the full technical report for DeepSeek-V4 on April 24, 2026, titled 'DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence.' The paper details FP4 quantization-awa
arXiv:2605.04759v1 Announce Type: new Abstract: Transformer based pre-trained large language models have become ubiquitous. There is increasing evidence to suggest that even with large scale pre-train
arXiv:2605.04752v1 Announce Type: new Abstract: Accurate traffic congestion classification requires models that jointly capture roadway scene context and non-stationary traffic motion, yet most prior
arXiv:2603.01097v2 Announce Type: replace Abstract: Continuous knowledge updating for pre-trained large language models (LLMs) is increasingly necessary yet remains challenging. Although inference-tim
arXiv:2510.12061v2 Announce Type: replace Abstract: Effective disaster response is essential for safeguarding lives and property. Existing statistical approaches often lack semantic context, generaliz
arXiv:2605.03160v1 Announce Type: new Abstract: The standard sparse-autoencoder (SAE) interpretability protocol labels each feature from its top-activating contexts and validates by single-feature ste
arXiv:2605.03720v1 Announce Type: new Abstract: Recent advances in Large Reasoning Models (LRMs) trained with Long Chain-of-Thought have demonstrated remarkable capabilities in code generation and mat
arXiv:2602.22575v2 Announce Type: replace Abstract: Attention scales quadratically with sequence length, fundamentally limiting long-context inference. Existing block-granularity sparsification can re
Strong Opinions, Loosely Held on Agent + Harness Engineering: 1. You can outperform any default harness+model (including codex & claude code) on pretty much any Task by engineering the harness around
Welcome to The Blueprint, a new feature where we highlight how Google Cloud customers are tackling unique and common challenges across industries using the latest AI and cloud technologies. We hope to
arXiv:2509.24552v3 Announce Type: replace Abstract: Recent works show that hybrid architectures combining local sliding window attention layers and global attention layers outperform either of these a
arXiv:2605.01910v1 Announce Type: new Abstract: Autoregressive decoding becomes bandwidth-limited at long contexts, as generating each token requires reading all n_k key and value vectors from KV cach
arXiv:2605.01399v1 Announce Type: new Abstract: The conventional Retrieval-Augmented Generation (RAG) paradigm of injecting raw retrieved texts into the Large Language Model (LLM)'s context often resu
arXiv:2603.15949v3 Announce Type: replace Abstract: Large Language Models have demonstrated strong multilingual fluency, yet fluency alone does not guarantee socially appropriate language use. In high
arXiv:2407.10853v5 Announce Type: replace Abstract: Bias and fairness risks in Large Language Models (LLMs) vary substantially across deployment contexts, yet existing approaches lack systematic guida
arXiv:2605.00298v1 Announce Type: new Abstract: Deploying reinforcement learning policies in the real world requires adapting to time-varying environments. We study this problem in the contextual Mark
Aura is a local-first AI daemon that enhances tools with persistent memory capabilities, claim verification features, and Model Context Protocol (MCP) observability. The project appears designed to ru
arXiv:2601.01885v2 Announce Type: replace Abstract: Large language model (LLM) agents face fundamental limitations in long-horizon reasoning due to finite context windows, making effective memory mana
arXiv:2604.28181v1 Announce Type: new Abstract: Realistic long-horizon productivity work is strongly conditioned on user-specific computer environments, where much of the work context is stored and or
arXiv:2604.25975v1 Announce Type: cross Abstract: Key-value (KV) caching is essential for large language model inference, yet its memory overhead poses a critical bottleneck for long-context generatio
arXiv:2511.07689v2 Announce Type: replace-cross Abstract: Evaluating the factual consistency of abstractive text summarization remains a significant challenge, particularly for long documents, where c
arXiv:2604.26951v1 Announce Type: cross Abstract: Diffusion large language models (dLLMs) offer parallel decoding and bidirectional context, but state-of-the-art dLLMs require billions of parameters f
arXiv:2512.12087v3 Announce Type: replace Abstract: The growing demand for long-context inference capabilities in Large Language Models (LLMs) has intensified the computational and memory bottlenecks
arXiv:2604.24971v1 Announce Type: cross Abstract: We present PolyKV, a system in which multiple concurrent inference agents share a single, asymmetrically compressed KV cache pool. Rather than allocat
The momentum is undeniable: the world’s fastest-growing AI startups are building with Google Cloud. Instead of stitching together fragmented point solutions, founders are building their businesses her
arXiv:2604.23134v1 Announce Type: new Abstract: Accurate molecular representations are critical for drug discovery, and a central challenge lies in capturing the chemical environment of molecular frag
arXiv:2505.11334v4 Announce Type: replace Abstract: This work aims at a challenging task: human action-reaction synthesis, i.e., generating human reactions conditioned on the action sequence of anothe
arXiv:2604.24187v1 Announce Type: new Abstract: Wide Field-of-View (WFoV) reconstruction enhances 3D ultrasound imaging by providing valuable anatomical context for segmentation models and visualizati
arXiv:2604.22782v1 Announce Type: cross Abstract: Serving transformer language models with high throughput requires caching Key-Values (KVs) to avoid redundant computation during autoregressive genera
arXiv:2603.03456v2 Announce Type: replace Abstract: Coding agents are increasingly deployed autonomously, at scale, and over long-context horizons. To be effective and safe, these agents must navigate
arXiv:2604.22661v1 Announce Type: cross Abstract: Large Language Models (LLMs) have made query reformulation ubiquitous in modern retrieval and Retrieval-Augmented Generation (RAG) pipelines, enabling
arXiv:2604.22050v1 Announce Type: cross Abstract: Transformers are mostly relying on softmax attention, which introduces quadratic complexity with respect to sequence length and remains a major bottle
In this post, we show how connecting the Visier Workforce AI platform with Amazon Quick through Model Context Protocol (MCP) gives every knowledge worker a unified agentic workspace to ask questions i
arXiv:2603.26791v2 Announce Type: replace-cross Abstract: Assessing a cited paper's impact is typically done by analyzing its citation context in isolation within the citing paper. While this focuses
Chinese AI lab DeepSeek's last model release was V3.2 (and V3.2 Speciale) last December. They just dropped the first of their hotly anticipated V4 series in the shape of two preview models, DeepSeek-V
arXiv:2603.22823v3 Announce Type: replace Abstract: Context. The problem of comparative evaluation of communication protocols for task orchestration by large language model (LLM) agents is considered.
arXiv:2507.03933v3 Announce Type: replace Abstract: Multilingual Large Language Models considerably changed how technologies influence language. While previous technologies could mediate or assist hum
arXiv:2507.01829v2 Announce Type: replace-cross Abstract: Multi-timescale sequence modeling relies on capturing both local fast dynamics and global slow context; yet, maintaining these capabilities un
arXiv:2604.21896v1 Announce Type: new Abstract: This paper introduces a new paradigm for AI game programming, leveraging large language models (LLMs) to extend and operationalize Claude Shannon's taxo
arXiv:2604.20904v1 Announce Type: cross Abstract: Information handling practices of LLM agents are broadly misaligned with the contextual privacy expectations of their users. Contextual Integrity (CI)
arXiv:2604.21893v1 Announce Type: cross Abstract: Geographic context is often consider relevant to motor insurance risk, yet public actuarial datasets provide limited location identifiers, constrainin
arXiv:2510.26615v3 Announce Type: replace Abstract: Retrieval-augmented generation (RAG) extends large language models (LLMs) with external knowledge, but it must balance limited effective context, re
arXiv:2604.21139v1 Announce Type: new Abstract: Language models must bind entities to the attributes they possess and maintain several such binding relationships within a context. We study how multipl
arXiv:2604.20849v1 Announce Type: cross Abstract: Retrieval-augmented generation over semi-structured sources such as HTML is constrained by a mismatch between document structure and the flat, sequenc
arXiv:2511.02849v2 Announce Type: replace-cross Abstract: Individualized therapy is driven forward by medical data analysis, which provides insight into the patient's context. In particular, for Type
arXiv:2604.19921v1 Announce Type: new Abstract: Negation is a common and important semantic feature in natural language, yet Large Language Models (LLMs) struggle when negation is involved in natural
arXiv:2604.20516v1 Announce Type: cross Abstract: Determining identifiability of causal effects from observational data under latent confounding is a central challenge in causal inference. For linear
arXiv:2604.20487v1 Announce Type: cross Abstract: Large language models (LLMs) encode knowledge in parametric weights, making it costly to update or extend without retraining. Retrieval-augmented gene
arXiv:2604.20819v1 Announce Type: new Abstract: The scalability of long-context large language models is fundamentally limited by the quadratic memory cost of exact self-attention, which often leads t
arXiv:2604.19689v1 Announce Type: new Abstract: Understanding artworks requires multi-step reasoning over visual content and cultural, historical, and stylistic context. While recent multimodal large
arXiv:2604.18988v1 Announce Type: new Abstract: Multimodal empathetic response generation (MERG) aims to generate emotionally engaging and empathetic responses based on users' multimodal contexts. Exi
arXiv:2601.03512v2 Announce Type: replace-cross Abstract: Code translation across multiple programming languages is essential yet challenging due to two vital obstacles: scarcity of parallel data pair
arXiv:2604.19052v1 Announce Type: new Abstract: Understanding a discourse requires tracking entities and the relations that hold between them. While Large Language Models (LLMs) perform well on relati
arXiv:2604.19005v1 Announce Type: new Abstract: Half-truths, claims that are factually correct yet misleading due to omitted context, remain a blind spot for fact verification systems focused on expli