Knowledge catalogue
arXiv:2605.31518v1 Announce Type: new Abstract: Sparse autoencoders (SAEs) decompose neural network activations into interpretable features, but many learned features never activate, a problem called
arXiv:2605.30512v1 Announce Type: new Abstract: Generating physics diagrams from text requires strict adherence to physical laws. While current generative models produce visually plausible outputs, th
arXiv:2605.31558v1 Announce Type: cross Abstract: Transformer-based language models are widespread in today's society. As such, understanding the mechanisms by which they solve structured tasks and pr
arXiv:2511.04393v2 Announce Type: replace Abstract: Large language models (LLMs) are increasingly deployed as 'agents' for decision-making (DM) in interactive and dynamic environments. Yet, since they
arXiv:2605.30711v1 Announce Type: cross Abstract: Agentic LLMs must continuously decide whether newly extracted facts should be added, merged with existing memories, or ignored, yet prior work has foc
arXiv:2605.30837v1 Announce Type: cross Abstract: Prompt-injection detectors are heterogeneous: each is strong on a different slice of attacks, and none is always reliable. Yet existing systems still
arXiv:2605.31393v1 Announce Type: cross Abstract: Sign language translation (SLT) remains constrained by limited paired sign-video/text corpora and heavy-tailed target vocabularies. We study target-si
arXiv:2605.30454v1 Announce Type: cross Abstract: Tool-augmented LLM agents are vulnerable to prompt injection: a third party who controls part of the agent's context can plant instructions that the a
arXiv:2605.31452v1 Announce Type: new Abstract: Building on our previous work, this paper develops practical, low-barrier methods for freelance translators and smaller language service providers to ev
arXiv:2605.29741v1 Announce Type: new Abstract: The dominance of colonial languages in African education and scientific communication limits how hundreds of millions of speakers of African languages a
arXiv:2605.29801v1 Announce Type: new Abstract: Modern open-world agents such as OpenClaw exhibit powerful cross-environment execution capabilities yet introduce broad new safety risk sources. Meanwhi
arXiv:2605.29434v1 Announce Type: cross Abstract: Existing sentence-level watermarking methods enhance robustness to paraphrasing by anchoring watermarks in sentence semantics. However, their prefix-b
arXiv:2605.29664v1 Announce Type: cross Abstract: Pipeline parallelism is essential for large-scale model training, but existing asynchronous approaches often degrade convergence due to parameter mism
Another proof point for the open-weights thesis. From @RampLabs: 'If we built this again, we'd lean more on open-weight models.' Ramp pointed 10K agents at their own backend. Kimi K2.6 and DeepSeek V4
arXiv:2605.29139v1 Announce Type: cross Abstract: Federated Conformal RAG (FC-RAG) provides distribution-free coverage for a bandwidth-limited swarm of weak language models, but only at a fixed horizo
arXiv:2605.29400v1 Announce Type: new Abstract: We benchmark three supervised fine-tuned models against frontier zero-shot baselines on a 661-row held-out slice of PiSAR (Persona, intent, Screen, Acti
arXiv:2410.10398v3 Announce Type: replace-cross Abstract: As large language models (LLMs) increasingly engage in complex social interactions, ensuring that their behaviors align with human ethical pri
arXiv:2602.04729v2 Announce Type: replace Abstract: We present a large-scale human evaluation benchmark for assessing cultural localisation in machine translation produced by state-of-the-art multilin
arXiv:2605.28830v1 Announce Type: cross Abstract: As Large Language Models (LLMs) are increasingly deployed in safety-critical applications, robust content moderation becomes essential. We present a c
arXiv:2605.29225v1 Announce Type: new Abstract: Self-evolving agents improve over time by reflecting on past failures, but existing evaluation is limited in two ways: it measures only task scores, lea
arXiv:2605.29365v1 Announce Type: new Abstract: Formality transfer is commonly framed as a symmetric bidirectional task between informal and formal registers. We argue that this framing conceals a sup
Claude really can roleplay an economist. I love this little comment Claude made after some robustness checks on the paper it wrote: 'On a 1–10 identification scale, I'd now put the paper at about 4.5
arXiv:2604.06805v2 Announce Type: replace Abstract: Multi-step Chain-of-Thought (CoT) has significantly advanced the mathematical reasoning capabilities of LLMs by leveraging explicit reasoning steps.
arXiv:2605.29446v1 Announce Type: new Abstract: Miller-index identification from powder XRD patterns requires capabilities untested by existing multimodal benchmarks: the model must read a narrow peak
arXiv:2605.29897v1 Announce Type: new Abstract: Evaluating the cultural awareness of large language models is crucial to ensure the fairness of generated text and the generalizability of applications
arXiv:2605.29427v1 Announce Type: new Abstract: As large language models (LLMs) are increasingly deployed in financial services, a single non-compliant interaction can expose institutions to regulator
I’m going to let you in on a secret. I made a mistake once.* In August 2024.* It’s true. I actually got something wrong. I predicted that there would be a collapse of the AI bubble (which in my judgem
arXiv:2605.29473v1 Announce Type: cross Abstract: Language models are increasingly being deployed for conversational support in informal caregiving contexts, where interactions often extend beyond inf
arXiv:2605.29459v1 Announce Type: new Abstract: Large language models route every input through a learned embedding table of shape |V| x d_model, consuming hundreds of millions to billions of trainabl
arXiv:2605.29928v1 Announce Type: cross Abstract: As AI-generated and AI-assisted content floods online spaces, source labels attached to such content can distort human reasoning judgments, with downs
arXiv:2605.30273v1 Announce Type: cross Abstract: Large language models (LLMs) show promise in generating supportive responses for mental health queries, but improving their usefulness, empathy, and s
arXiv:2602.14399v2 Announce Type: replace Abstract: Multi-turn jailbreak attacks have proven effective against text-only large language models (LLMs), where malicious content is gradually introduced t
arXiv:2605.29563v1 Announce Type: new Abstract: Can VLMs predict how each camera move changes the view, and plan many such moves ahead? We call this capability view planning, requiring (1)understandin
arXiv:2605.30094v1 Announce Type: new Abstract: Poker is a landmark challenge for artificial intelligence. The dominant approach relies on equilibrium solvers built on counterfactual regret minimizati
arXiv:2603.05488v4 Announce Type: replace-cross Abstract: We provide evidence of performative chain-of-thought (CoT) in reasoning models, where a model becomes strongly confident in its final answer,
arXiv:2509.23694v5 Announce Type: replace Abstract: Search agents connect LLMs to the Internet, enabling them to access broader and more up-to-date information. However, this also introduces a new thr
arXiv:2605.29059v1 Announce Type: cross Abstract: Smart contract decompilation aims to recover high-level source code from bytecode, but evaluating decompilers remains difficult because existing studi
arXiv:2605.06322v2 Announce Type: replace Abstract: Language models for molecular design have scaled to hundreds of millions of parameters, yet how they learn chemical grammar is poorly understood. We
arXiv:2605.29000v1 Announce Type: new Abstract: Traditional lossless text compression preserves every byte, but its gains on natural language are often modest in realistic operating regimes. We study
arXiv:2603.23971v2 Announce Type: replace-cross Abstract: Developers and consumers increasingly choose reasoning models (RMs) based on their listed API prices. However, how accurately do these prices
arXiv:2605.29096v1 Announce Type: new Abstract: This paper examines records retrieved from the ClinicalTrials.gov registry to characterize temporal trends in AI terminology and the geographical distri
We’re taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders develop new biodefense and pandemic preparedness capabilities. - Expanding trus
arXiv:2605.27485v1 Announce Type: cross Abstract: Formal verification offers a path to provably correct software, but writing verified code remains expensive enough that the technique is rarely used i
arXiv:2605.28655v1 Announce Type: new Abstract: Scientific research proceeds through iterative cycles of hypothesis generation, experiment design, execution, and revision. AI agents can automate parts
arXiv:2605.27700v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly used to generate scientific reports, but they can produce references that appear plausible while contain
arXiv:2603.02097v5 Announce Type: replace Abstract: Open-ended medical LLM evaluation remains weakly grounded in physician-calibrated coverage of clinically relevant response criteria, especially in l
arXiv:2603.24631v2 Announce Type: replace-cross Abstract: Code agents resolve 65-70% of SWE-bench Verified issues, but Pass@1 cannot tell us why the rest fail, and, as we show, capable-model failures
arXiv:2605.27654v1 Announce Type: cross Abstract: Generative translation systems are cultural technologies because they decide how socially meaningful cues are rendered within culturally specific gram
arXiv:2512.00349v2 Announce Type: replace Abstract: Are frontier AI systems becoming more capable? Certainly. Yet such progress is not an unalloyed blessing but rather a Trojan horse: behind their per
arXiv:2605.27858v1 Announce Type: cross Abstract: Claim verification splits between end-to-end classifiers that are accurate but yields no inspectable traces, and decomposition-based methods produce i
arXiv:2605.27773v1 Announce Type: cross Abstract: When a language model sees a document contradicting its training knowledge, it must choose: follow the document or trust itself. Prior work proved thi
arXiv:2605.27387v1 Announce Type: cross Abstract: Diffusion models promise efficient parallel text generation but rely on bidirectional attention, creating a structural mismatch with pre-trained Autor
arXiv:2605.27595v1 Announce Type: cross Abstract: Large Language Models (LLMs) are being rapidly adopted in agricultural imaging applications, ranging from crop interpretation to synthetic field image
arXiv:2605.27646v1 Announce Type: cross Abstract: We propose extbf{Hurwitz Quaternion Multiplicative Quantization (HQMQ)}, a extbf{calibration-free} method for KV cache compression of large language m
arXiv:2605.27375v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly acting as autonomous agents, but their continuous interaction with the environment can lead to in-context
arXiv:2304.12986v3 Announce Type: replace-cross Abstract: The development of large-scale Chinese language models is flourishing, yet there is a lack of corresponding capability assessments. Therefore,
arXiv:2605.27853v1 Announce Type: new Abstract: We present MolLingo, a multi-agent system that emulates the reasoning process of a chemist to automate molecular design. Existing LLM-based approaches e
arXiv:2605.27545v1 Announce Type: new Abstract: Jailbreak attacks on multimodal AI systems remain underexplored, even though unsafe image generation can have more severe consequences than unsafe text
arXiv:2605.27980v1 Announce Type: cross Abstract: The ability to process ultra-long contexts is crucial for large language models (LLMs) to perform long-horizon tasks. While recent efforts have extend
arXiv:2605.28207v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) is now the dominant architecture for frontier language models, yet it requires all expert parameters to be loaded in memory,