Knowledge catalogue
On my two DGX Sparks, Swift 1.5 (UkisAI's reasoning-efficient fine-tune of Qwen3.8-Flash-Next) running on veloGB10 (https://github.com/sf-stav/veloGB10, sf-stav's Rust/CUDA engine built only for GB10)
arXiv:2610.03565v1 Announce Type: new Abstract: We describe the Writerslogic systems for three PAN at CLEF 2026 shared tasks (Reasoning Trajectory Detection, Voight-Kampff Generative AI Detection, and
I use ChatGPT a lot. Probably a lot more than is reasonable. I read David Robinson’s Atlantic piece today about leaving OpenAI and his criticism of its culture of trial-and-error / “iterative deployme
arXiv:2610.01710v1 Announce Type: new Abstract: Visual grounding localizes an object described by language with a bounding box. Most multimodal grounding models compress target identification, spatial
arXiv:2610.02005v1 Announce Type: new Abstract: A multi-LLM council lets several large language models (LLMs) deliberate on a question and return an answer together with a confidence estimate. As thes
arXiv:2610.00791v1 Announce Type: new Abstract: Enterprise information is represented through artifacts shaped by applications, projects, technologies, organizational boundaries, and local requirement
arXiv:2607.16956v2 Announce Type: replace Abstract: Social navigation requires the robot to reason and respond in complex real-world environments. While recent works attempt to incorporate human-level
Hi all! I am a former SWE who has recently transitioned into inference engineering. I launched a side hustle a few months back and I've just taken it full time due to excessive demand. Its been such a
arXiv:2610.00437v1 Announce Type: new Abstract: LLM agents generate intermediate reasoning and actions token by token, making extended interactions slow and computationally expensive. Jev-style models
arXiv:2610.01569v1 Announce Type: cross Abstract: Unmanned aerial vehicle (UAV) swarms increasingly rely on language-model agents to provide adaptive mission-level reasoning in uncertain environments.
arXiv:2610.01383v1 Announce Type: new Abstract: Analogical reasoning involves identifying and preserving relational structures across domains. However, existing approaches to AI-driven multimodal anal
arXiv:2608.28421v2 Announce Type: replace Abstract: Post training a language model to reason means updating its weights. Supervised finetuning and reinforcement learning both place the acquired capabi
arXiv:2508.08882v5 Announce Type: replace Abstract: Recent advances in multi-agent systems highlight the potential of specialized small agents that collaborate via division of labor. Existing tool-int
Ten years ago, AlphaGo’s Move 37 shocked the world. It wasn’t intuition alone that produced it. AlphaGo could search possible futures, test its instincts and reason about what would happen next. In a
arXiv:2610.02054v1 Announce Type: new Abstract: Vision-language-action models benefit from the understanding and reasoning capabilities of pretrained vision-language models, but action-only supervisio
arXiv:2610.00757v1 Announce Type: new Abstract: Long-video question answering is limited by the high cost of visual tokens and by the fixed context width of current VLMs. A long-video question may req
arXiv:2609.39277v1 Announce Type: cross Abstract: Looped models reason by applying the same block of weights many times, so compressing that block saves memory traffic on every loop. Compressed looped
arXiv:2609.39687v1 Announce Type: new Abstract: On-policy self-distillation (OPSD) trains mathematical reasoning models using a privileged teacher that sees a reference solution and supervises student
arXiv:2609.40124v1 Announce Type: new Abstract: Bias has long been studied in social psychology and cognitive science, where decades of research have produced a body of validated interventions that re
arXiv:2609.40306v1 Announce Type: cross Abstract: Pretrained robot policies provide useful action priors, but long-horizon manipulation still requires coordination between semantic reasoning and physi
arXiv:2609.38805v1 Announce Type: cross Abstract: LLM agents often admit multiple high-quality solutions to the same task, differing in reasoning structure, tool-use pattern, or interaction trajectory
arXiv:2510.17932v5 Announce Type: replace-cross Abstract: We introduce Chart2Code, a new benchmark for evaluating the chart understanding and code generation capabilities of large multimodal models (L
arXiv:2609.39321v1 Announce Type: cross Abstract: Group Relative Policy Optimization (GRPO) has emerged as a memory-efficient reinforcement fine-tuning (RFT) technique for reasoning-intensive tasks. H
arXiv:2609.39588v1 Announce Type: new Abstract: We push the frontier of large-scale spatial intelligence in Vision-Language Models (VLMs) and introduce the first benchmark that probes geographical lay
arXiv:2604.08454v2 Announce Type: replace Abstract: Large language models are increasingly deployed in high-stakes domains, where confident yet incorrect inferences may cause severe real-world harm, b
arXiv:2609.39146v1 Announce Type: new Abstract: Multi-agent debate (MAD) can improve large language model (LLM) reasoning by allowing multiple agents to exchange and critique their answers to the same
arXiv:2609.39076v1 Announce Type: new Abstract: A pool of language models can collaborate and improve collectively by learning from one another's responses. These interactions depend on the instructio
arXiv:2609.39334v1 Announce Type: cross Abstract: Test-time scaling has recently emerged as a powerful approach for improving LLM reasoning by allocating additional computation during inference, subst
arXiv:2609.40115v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has become an important approach for improving reasoning during post-training. Recent work suggest
arXiv:2609.38964v1 Announce Type: new Abstract: Multi-agent debate (MAD) is often used to improve large language model (LLM) reasoning, but sequential debate is rarely a neutral aggregator of agents'
arXiv:2603.14733v2 Announce Type: replace Abstract: Multimodal Large Language Models have achieved strong performance in single-video understanding, yet their ability to reason across multiple videos
arXiv:2609.36985v1 Announce Type: cross Abstract: The central challenge of world modeling is to learn representations that capture how the world evolves. However, existing world models predominantly r
arXiv:2609.37510v1 Announce Type: new Abstract: On-policy distillation (OPD) trains a student on its own reasoning trajectories using feedback from a stronger teacher. Teacher interventions can improv
arXiv:2609.38027v1 Announce Type: new Abstract: In recent years, the performance of large language models (LLMs) on reasoning tasks has been remarkable, even surpassing human capabilities on various b
arXiv:2609.37044v1 Announce Type: new Abstract: Explicit intermediate reasoning gives large language models (LLMs) a stronger problem-solving mode. We study learning from this think-mode advantage via
arXiv:2609.37825v1 Announce Type: cross Abstract: Assigning credit to intermediate steps remains a central challenge in training Large Language Models (LLMs) on multi-step reasoning tasks with sparse
arXiv:2609.36813v1 Announce Type: new Abstract: Large language models (LLMs) exhibit strong general capabilities that mechanistic interpretability has attributed to sparse computational circuits. Howe
arXiv:2602.08847v2 Announce Type: replace Abstract: Multi-agent LLM systems enable advanced reasoning and tool use via role specialization, yet reliable reinforcement learning (RL) post-training for s
IQuest-Q1 is a Mixture-of-Experts (MoE) model developed by IQuest for agentic coding, reasoning, and multi-step tool use. It comprises approximately 320B total parameters, with an estimated 15B parame
arXiv:2609.34491v1 Announce Type: new Abstract: Small-molecule optimization integrates medicinal-chemistry reasoning and computational evidence through iterative, multi-objective decisions. When large
arXiv:2609.31947v1 Announce Type: new Abstract: Hybrid transformer architectures that replace most softmax attention layers with linear attention offer transformer-level quality at a fraction of the m
arXiv:2609.34806v1 Announce Type: cross Abstract: Reasoning about temporal structure of audio recordings requires Large Audio Language Models (LALMs) to associate sound events with their temporal posi
Thanks to tcclaviger, vLLM now has expert RAM offloading support (link). This makes frontier models much more accessible on a local setup! I was able to run the original DeepSeek-V4-Flash-Vision-Exp o
arXiv:2609.34905v1 Announce Type: new Abstract: Power sampling has emerged as a training-free approach to LLM reasoning, eliciting capabilities comparable to reinforcement learning by sharpening the m
arXiv:2609.34044v1 Announce Type: new Abstract: Reasoning vision-language models (VLMs) process images and videos as long sequences of visual tokens, making inference expensive. Training-free token pr
arXiv:2609.33780v1 Announce Type: new Abstract: Verified solutions are not equally useful for preparing reasoning models for reinforcement learning (RL). We present a comprehensive study of route dive
arXiv:2609.30484v1 Announce Type: new Abstract: While large language models (LLMs) have achieved remarkable linguistic capabilities, a profound question lingers at their core: do these models truly co
arXiv:2609.31186v1 Announce Type: new Abstract: Artificial intelligence is advancing rapidly, with increasingly capable systems taking larger roles in reasoning, decision-making, scientific discovery,
arXiv:2609.31029v1 Announce Type: new Abstract: Reasoning systems usually treat premise use as a question of relevance: if a fact is available and useful, it may be selected for inference. Authorizati
arXiv:2609.31382v1 Announce Type: cross Abstract: Long-context understanding requires large language models (LLMs) to reason over lengthy documents, conversations, and code, yet task-relevant evidence
arXiv:2609.30328v1 Announce Type: new Abstract: When one language model judges whether another's code is correct, it does not report the absence of evidence. It returns a confident verdict with reason
On Friday I gave the closing keynote at the WeAreDevelopers World Congress North America in San Jose. I tied together the key trends from the past year into a chronological exploration of everything t
I posted an updated GPT-OSS template a couple of months ago, which was based on Unsloth's version. It turns out that both Unsloth's version (and, thus, mine) contain a very serious bug that can degrad
arXiv:2609.29861v1 Announce Type: new Abstract: We investigate whether GPT-6-Astra, a general-purpose foundation model, can navigate unfamiliar environments using its own perception, reasoning, and de
arXiv:2609.29166v1 Announce Type: cross Abstract: Physical AI aims to build embodied agents that perceive the world, understand and reason about it, and decide how to act. Yet the field has focused pr
Since my last post, I've been thinking about different options for dynamic performance degradation, trying to squeeze as much high-quality inference out of my GPU as I can. Over the weekend I read thi
arXiv:2609.27572v1 Announce Type: cross Abstract: Reinforcement learning (RL) has emerged as a key paradigm for improving the reasoning capabilities of large language models (LLMs). However, existing
arXiv:2609.27150v1 Announce Type: new Abstract: Multi-agent debate (MAD) has emerged as a promising paradigm for improving the reasoning accuracy of large language models (LLMs) through iterative peer
arXiv:2607.10128v3 Announce Type: replace Abstract: Recursive models show promise on reasoning and language tasks, yet their test-time scaling lacks a principled criterion for selecting trajectories o
Hey everyone, Jovan from UkisAI here! Today, we are introducing Swift, a family of efficient reasoning LLMs based on Qwen, trained by penalizing tokens related to pathological overthinking patterns an