Knowledge catalogue
arXiv:2505.17086v4 Announce Type: replace Abstract: Large Language Models (LLMs) equipped with modern Retrieval-Augmented Generation (RAG) systems often employ multi-turn interaction pipelines to inte
arXiv:2602.14200v4 Announce Type: replace Abstract: Time Series Language Models (TSLMs) are emerging as unified models for reasoning over continuous signals in natural language. However, long-context
arXiv:2604.09408v1 Announce Type: new Abstract: Frontier coding agents solve complex tasks when given complete context but collapse when specifications are incomplete or ambiguous. The bottleneck is n
arXiv:2610.20419v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) models have become almost a cornerstone of open-domain question answering tasks; in which the quality of answers ge
As enterprises shift their AI usage from experimental chatbots to autonomous agents, teams are hitting a critical bottleneck: accessing and reasoning over data scattered across multiple clouds, SaaS p
arXiv:2610.15642v1 Announce Type: new Abstract: Under a limited context budget, a long-term agent must recall not merely memories that are individually relevant, but evidence that is collectively suff
arXiv:2603.09200v2 Announce Type: replace-cross Abstract: Situational awareness, the capacity of an AI system to recognize its own nature, understand its training and deployment context, and reason st
arXiv:2610.19814v1 Announce Type: new Abstract: Transformer-based tabular foundation models (TFMs) achieve state-of-the-art performance on small to medium-sized tabular prediction tasks while processi
arXiv:2610.14688v1 Announce Type: new Abstract: LLM-augmented data workflows often involve multiple heterogeneous models operating over substantial shared context, such as schemas, tool specifications
arXiv:2603.29602v2 Announce Type: replace-cross Abstract: Existing multi-turn image editing paradigms are often confined to isolated single-step execution. Due to a lack of context-awareness and close
arXiv:2610.14219v1 Announce Type: new Abstract: Softmax attention in Transformers suffers from quadratic complexity in sequence length, making it impractical for long-context applications. Linear atte
A coding agent finishes a change and wants to open a GitHub issue with a link to its results. The agent can describe the issue, but creating it belongs to a different system. It needs a tool that can
arXiv:2610.13708v1 Announce Type: new Abstract: Long-context serving runs out of KV memory before the hardware reaches its useful throughput. HiSparse stores history in host memory and fetches only th
arXiv:2610.13494v1 Announce Type: new Abstract: Existing LLM-based agent approaches for long-horizon machine learning engineering (MLE) often use context engineering, implemented through custom harnes
I get a lot of hate whenever I tell people I run a Q4 quant of Qwen3.8 27B with 262k KV 80t/s with vision, all while running 2 screens on windows on the same card. People seem to think I am lying or m
arXiv:2610.13202v1 Announce Type: cross Abstract: Fast rollout generation for long-context reinforcement learning with verifiable rewards (RLVR) requires reducing KV-cache traffic without changing the
arXiv:2610.13325v1 Announce Type: cross Abstract: Agentic workloads resend long contexts that barely change between calls. Prefix caching now skips most of their prefill, yet every request must still
arXiv:2610.13237v1 Announce Type: cross Abstract: Joint-embedding predictive architectures (JEPAs) learn by predicting teacher representations of masked regions from visible context. Point prediction
arXiv:2610.13695v1 Announce Type: new Abstract: Any automation within reliability-critical contexts such as medicine or law has to be transparent and verifiable. Despite their advantages, current LLM-
Based on my experiments and feedback from claude, it seems extremely unlikely and even if I get it working, not practical because of minimum context. But just asking here as a last effort. I'm able to
arXiv:2610.12133v1 Announce Type: new Abstract: Many key-value (KV) caches are compressed before anyone knows what will be asked of them: a document cached for retrieval, a prompt prefix shared across
arXiv:2610.01108v2 Announce Type: replace Abstract: Retrieval-based speculative decoding (SD) drafts tokens by copying continuations from existing text, which suits coding agents that repeatedly repro
arXiv:2610.06919v1 Announce Type: new Abstract: This paper concerns how semantic context determines geometry in learned vector representations. Similarity is typically measured using cosine similarity
Industry-specific AI brings business context to agents handling enterprise workflows. Building around industry processes can help companies address reliability concerns as they automate more tasks. In
arXiv:2610.08630v1 Announce Type: new Abstract: Recently Large Language Models (LLMs) and LLM-based agents increasingly need to incorporate knowledge acquired after pretraining, e.g., domain facts, us
arXiv:2610.06902v1 Announce Type: new Abstract: Retrieval-augmented generation grounds language models in external context, but for long documents flat top-k retrieval can cluster on a single region a
AI agent memory is creating new demands on infrastructure as agents run longer sessions and spread across the enterprise. Retaining that context and making it available when needed puts pressure on me
arXiv:2610.02772v1 Announce Type: cross Abstract: Large language models (LLMs) increasingly serve as general-purpose interfaces to factual knowledge, but their parameters do not automatically reflect
arXiv:2610.02815v1 Announce Type: new Abstract: Long chain-of-thought reasoning substantially increases KV-cache memory during autoregressive decoding, as every generated token introduces new key and
I've been wanting to test out LLM inference on FPGAs for a while now, but didn't have a big reason too since there were no frontier class models at 9B-27B scale (3.6 27B was great still but it didn't
arXiv:2610.01415v1 Announce Type: new Abstract: Large language model (LLM) agents can now undertake increasingly complex tasks, but the way they organize interaction history into memory does not ensur
arXiv:2610.01230v1 Announce Type: new Abstract: Efficient long-context inference is essential for large language models (LLMs), yet it poses a severe computational bottleneck. Hash-based retrieval off
arXiv:2610.02153v1 Announce Type: new Abstract: Long-horizon autoregressive video generation is limited by a finite context window. When an object or scene falls out of context, its fine-grained visua
arXiv:2610.01046v1 Announce Type: cross Abstract: Collaboration depends on shared context, and technical documentation is one way that context persists across people and AI teammates. Specificity, the
I have been building a somewhat unusual local inference machine around two Huawei Atlas 300I Duo cards. They are relatively inexpensive, passive, dual-accelerator PCIe cards with 96 GB of device memor
arXiv:2609.34211v2 Announce Type: replace Abstract: In financial fraud detection, rich semantic context can provide important evidence for transaction behavior modeling and fraud reasoning. However, p
arXiv:2609.38660v1 Announce Type: cross Abstract: Long-form subtitle translation requires reasoning over discourse and cultural context spanning episodes or entire series, while maintaining consistent
arXiv:2609.38912v1 Announce Type: new Abstract: Agent harnesses govern how large language models (LLMs) gather context, invoke tools, verify results, preserve state, and terminate, largely affecting a
arXiv:2609.39494v1 Announce Type: new Abstract: Self-distillation with privileged context adapts a language model from demonstrations by letting the model, once conditioned on a reference response, te
arXiv:2609.38744v1 Announce Type: new Abstract: Predicting molecular properties for compounds that differ structurally from labeled training molecules is important for drug discovery and materials des
arXiv:2609.38840v1 Announce Type: cross Abstract: Single-cell RNA-seq representation learning is fundamentally label-free: cell identities, states, and contexts are not fixed training targets, so what
arXiv:2609.38712v1 Announce Type: new Abstract: Long-horizon agentic workflows require models to sustain repeated state-dependent actions all while the context grows, sub-task complexity changes, and
arXiv:2609.39661v1 Announce Type: new Abstract: Self-attention gives LLMs fine-grained, query-dependent access to context, but dense token interactions incur quadratic prefill cost and a key--value ca
arXiv:2609.36883v1 Announce Type: new Abstract: Tabular foundation models (TFMs) are increasingly popular because they deliver strong predictions on new datasets through in-context learning, without t
arXiv:2609.36722v1 Announce Type: new Abstract: Large language model (LLM) agents repeatedly load reusable content, such as skills, documents, and memory entries, into the current context. Re-encoding
arXiv:2609.36364v1 Announce Type: cross Abstract: Standard video generators do not natively compact historical context into reusable memory tokens. As generation continues, the growing history makes i
arXiv:2609.36576v1 Announce Type: new Abstract: Agentic systems are now being widely used to orchestrate tools and reason over long contexts. However, the improving capabilities of the large language
arXiv:2609.37209v1 Announce Type: new Abstract: Pairwise losses are increasingly used for reward learning even when pointwise rewards are observed, with mixed empirical results. When and why do pairwi
arXiv:2609.37879v1 Announce Type: new Abstract: How many tokens from its context does a language model actually use, and what determines that number? We study this question through self-attention. Wit
arXiv:2609.38121v1 Announce Type: new Abstract: KV cache memory and bandwidth costs grow with context length and batch size, which limits efficient long-context inference. To address this bottleneck,
arXiv:2609.33746v1 Announce Type: new Abstract: At each decoding step a language model attends over the key-value (KV) cache of every earlier token, so at long context the attention call is bounded by
arXiv:2609.35497v1 Announce Type: new Abstract: Long video understanding relies on video memory to overcome the context limits of multimodal large language models. Existing methods follow a build-then
arXiv:2609.32898v1 Announce Type: new Abstract: Pretrained tabular foundation models process every dataset at a fixed depth, with inference costs growing with dataset size. To address this, we present
arXiv:2609.30288v1 Announce Type: new Abstract: In Transformer-based masked language models, attention is the primary mechanism for context mixing, but there are other ways to mix data across tokens.
arXiv:2609.31261v1 Announce Type: cross Abstract: The quadratic complexity of dense self-attention remains a central bottleneck for long-context language modeling. Many efficient alternatives address
arXiv:2609.31167v1 Announce Type: new Abstract: EEG foundation models increasingly use masked prediction to learn from unlabeled recordings, but optimizing this objective does not ensure transferable
I’m having a hard time finding out which one gives you fastest speed, maximum context with best possible quality. I can run unsloth qwen3.8 27b iq4_xs with 65k q8 kv, context without MTP and vision of
I am running qwen3.8 27b on my dual rtx 3090 (fp8 quant, unquantized cache, 129k context) and I think it works decently well with hermes, opencode etc. But! I am tempted by the new models coming out s
Running qwen3.8 27b nvfp4 on vllm at max context only gives around 8 agents with 32k context each. That doesnt seem like much; what use cases do people use multi-agent frameworks and find it helpful f
arXiv:2609.27925v1 Announce Type: new Abstract: Language models can over-condition on irrelevant preceding text: predictions already supported by local context may still change when distant, unrelated