Search: “context”

Benchmarked Muse Glimmer 30B on my RTX 5090 (32GB), 262k context, UD-Q5_K_M + dflash-kquant + mmproj. Workload Stock master + DFlash ngram-simple PR #26842 + DFlash Code patch 78 t/s 57 t/s 220-253 t/

arXiv:2608.07001v1 Announce Type: new Abstract: As large language models (LLMs) process increasingly long contexts, KV cache storage and repeated access have become a major bottleneck. Existing KV cac

arXiv:2608.07068v1 Announce Type: new Abstract: Long-horizon agents accumulate growing contexts during interaction, impairing performance and stability. Compact memory mitigates this problem by compre

*SupraLabs released a new experimental model!* SupraElegans-500K is a ~500,000-parameter causal language model built around a sparse, signed, recurrent neural graph. No Transformer, no attention mecha

I run the following on a 5090 and have been okay with its performance, it does most things somewhere 80-100 t/s, though that can slow down at full 262k context - more like 40 t/s at times. I use it pr

Open models give teams more room to run agent loops, with API prices at a fraction of GPT-5.6 Sol and Claude Fable 5. That matters as planning, tool calls, retries, and long contexts compound token us

arXiv:2608.04405v1 Announce Type: cross Abstract: Long-context large language models (LLMs) are increasingly deployed in real-world applications, yet self-attention remains a major efficiency bottlene

arXiv:2608.02947v1 Announce Type: cross Abstract: The attention score with rotary position embeddings (RoPE) decomposes exactly into a sum over its 2D-rotation frequency pairs, and each pair's wavelen

As you all know the model is 2.69B parameters with a 128K context window and purpose-built for multi-step agent workflows. What you are seeing is the Q4_K_M GGUF running on my own inference engine bui

In today’s retail environment, shoppers expect highly personalized product discovery experiences and conversational assistance that feels genuine, natural, and genuinely helpful. Today, successful pro

arXiv:2608.00036v1 Announce Type: new Abstract: Real-world document tasks often ask professionals to answer questions from annual reports, regulations, clinical guidelines, and technical manuals that

arXiv:2603.06274v2 Announce Type: replace-cross Abstract: The quadratic computational complexity of self-attention remains a fundamental bottleneck for scaling Large Language Models (LLMs) to long con

arXiv:2601.22546v2 Announce Type: replace-cross Abstract: The recent advancements in Large Language Models (LLMs) have attracted interest in exploring their in-context learning abilities and chain-of-

arXiv:2607.29032v1 Announce Type: cross Abstract: Large language model (LLM) agents increasingly operate over long interaction histories, where effective reasoning requires identifying and exploiting

what does the 'last human code review' look like? to @itamar_mar, ceo and founder of @QodoAI, it looks like two agents talking to each other, backed by a context engine specific to your organization.

arXiv:2607.27546v1 Announce Type: new Abstract: Tabular Foundation Models, such as TabPFN, have received a large amount of recent attention due to their performance on in-context tabular machine learn

arXiv:2607.25202v1 Announce Type: new Abstract: Conversational entrainment is well-studied in monolingual and written contexts, but remains underexplored in spoken code-switching (CSW). We present a n

arXiv:2607.25021v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) can connect visualization patterns to external causes, consequences, and domain knowledge, but the evidential b

arXiv:2607.25125v1 Announce Type: new Abstract: Despite rapid progress in Multi-modal Large Language Models (MLLMs), understanding long-form videos is still bottlenecked by limited context windows. Wh

arXiv:2607.24892v1 Announce Type: cross Abstract: Text-conditioned time-series forecasting predicts a series from both its numerical history and natural-language context, allowing forecasts to account

arXiv:2607.25357v1 Announce Type: cross Abstract: Long-context recall in linear-time sequence models highlights a tradeoff in how they write to memory. State-based linear models, such as state-space m

arXiv:2604.07263v2 Announce Type: replace-cross Abstract: Existing driving automation (DA) systems on production vehicles rely on human drivers to decide when to engage DA while requiring them to rema

Increasing the adoption of generative AI across the enterprise requires you to do more than deploy a generic chatbot with a custom wrapper. Interacting with business-critical databases demands absolut

arXiv:2607.24331v1 Announce Type: new Abstract: As the inference phase of Large Language Models (LLMs) requires handling long context windows, the Key-Value (KV) cache initially appears to address thi

arXiv:2607.24097v1 Announce Type: new Abstract: Memory-augmented LLM agents typically answer queries by retrieving relevant memories and feeding them directly to an answer model. This retrieval-as-evi

arXiv:2607.21752v1 Announce Type: new Abstract: Data-adaptive sparse attention masks substantially outperform fixed patterns (e.g., BigBird and Longformer) and can even exceed dense attention on long

arXiv:2601.08160v2 Announce Type: replace Abstract: Agentic memory systems have become critical for enabling LLM agents to maintain long-term context and retrieve relevant information efficiently. How

Running GLM-5.2 (the new glm-dsa arch), Unsloth UD-Q4_K_XL, on a 4-socket Xeon E7-8880 v4 box with 1TB RAM and a single RTX 3060 12GB. ik_llama.cpp, experts on CPU (--cpu-moe), 24 attention layers on

arXiv:2607.20827v1 Announce Type: new Abstract: LLM agents choose tools and arguments from context that mixes user requests, tool outputs, retrieved records, memory, and untrusted text. Evidence can b

arXiv:2607.21051v1 Announce Type: new Abstract: Real-world agent learning is often constrained by costly environment interactions, such as running time-consuming experiments or obtaining human feedbac

arXiv:2607.20484v1 Announce Type: new Abstract: Large Language Models (LLMs) are fundamentally limited by representation collapse, a bottleneck that severely degrades long-context performance. We iden

arXiv:2607.19363v1 Announce Type: new Abstract: Rotary Position Embedding (RoPE) is widely adopted in Transformers to encode positional information, yet standard implementations enforce a uniform freq

arXiv:2607.19082v1 Announce Type: cross Abstract: Autoregressive context models are foundational for learned image compression,but they suffer from slow serial inference. Existing acceleration methods

As adversarial AI threats accelerate attacks on code, security teams must counter them with machine-speed defenses that can automate code remediation and fight AI with AI. CodeMender is our managed co

arXiv:2607.11942v1 Announce Type: cross Abstract: KV-cache compression methods are predominantly evaluated with the query appended to the context before compression -- a query-aware protocol. Yet the

arXiv:2502.11554v3 Announce Type: replace-cross Abstract: Metaphors play a critical role in shaping user experiences with Voice User Interfaces (VUIs), yet existing designs often rely on static, human

arXiv:2607.07678v1 Announce Type: new Abstract: Rotary Position Embeddings (RoPE) provide transformers with a fixed grid of positional frequencies, yet trained models use these frequencies highly non-

arXiv:2607.05764v1 Announce Type: new Abstract: Answering questions over a set of transactional legal documents is most simply done by injecting the whole corpus into the LLM's context window on every

arXiv:2601.09448v3 Announce Type: replace-cross Abstract: Conventional audio equalization is a static process that requires manual and cumbersome adjustments to adapt to changing listening contexts (e

arXiv:2511.16893v3 Announce Type: replace Abstract: Specialized attention heads dubbed induction heads (IHs) have been argued to underlie the remarkable in-context learning capabilities of modern lang

arXiv:2607.05184v1 Announce Type: new Abstract: Self-distillation is a promising recipe for self-improvement in language models. In this setting, a model can serve as its own teacher when given privil

arXiv:2606.31650v1 Announce Type: cross Abstract: Long-horizon language agents must repeatedly interact with tools, accumulate evidence, and make decisions under bounded context windows. Existing cont

arXiv:2606.30317v1 Announce Type: cross Abstract: The Model Context Protocol (MCP), introduced by Anthropic in November 2024, defines a standardized interface for connecting large language models (LLM

arXiv:2506.12697v3 Announce Type: replace-cross Abstract: Small-object detection in Unmanned Aerial Vehicle (UAV) imagery requires preserving weak local evidence while using broader context to separat

arXiv:2606.30518v1 Announce Type: new Abstract: Retrieval-augmented generation (RAG) improves language models by grounding generation in external context. However, it can be fragile when the retrieved

arXiv:2506.00400v4 Announce Type: replace-cross Abstract: LLM-based prompt optimization, which uses LLM-provided ``textual gradients'' (feedback) to refine prompts, has emerged as an effective method