Every paper checked — nothing hidden

All papers

Every paper checked for brief #4 of 12–19 Oct 2026 — nothing is hidden. Brief papers first.

All17,313 In brief24 More (on-topic)3,333 Not relevant13,754 Not assessed87 Older352 Could not confirm93 Blogs with no paper266 All blogs284

Paper	Topic	Relevance
Adaptive and Proximal Federated Optimization for Transferable Network Intrusion Detection The authors study whether adaptive and proximal federated optimizers improve the detection of unseen attack classes in network intrusion detection. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on federated optimization for network intrusion detection, not AI model deployment safety.
DarKnight: Retrieval-Guided Agentic Framework for Breaking Dark Web CAPTCHA Gates This paper introduces DarKnight, a framework that uses locally hosted vision-language models to solve dark web CAPTCHAs during threat intelligence gathering. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Targets automated CAPTCHA cracking rather than AI model safety, guardrails, or robustness.
DarKnight: Retrieval-Guided Agentic Framework for Breaking Dark Web CAPTCHA Gates This paper introduces DarKnight, a framework that uses locally hosted vision-language models to solve dark web CAPTCHAs during threat intelligence gathering. Preprint	Computer vision for industrial applications via arxiv · 15 Oct, 15:06	Not relevant Solves CAPTCHAs for dark web crawling, not industrial manufacturing inspection.
Global and Local Entropy Tests for STL Files Under IEEE-754 Validity Constraints The authors derive statistical entropy tests under IEEE-754 validity constraints to detect non-randomness in encrypted stereolithography CAD files. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on cryptographic entropy statistics of 3D file formats, not AI systems.
Bilinear Compressive Security with Finitely Many Samples This paper investigates finite-sample key-recovery attacks against bilinear compressive security using a covariance-based estimation approach. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on mathematical cryptography for compressed sensing, not AI safety or deployment.
ESAU: Expeditious Swarm Remote Attestation for Pub-Sub IoT Networks with Untrusted Cloud-based Proxy Verifier This paper proposes ESAU, a lightweight swarm remote attestation framework for internet-of-things devices using untrusted cloud proxies. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Covers remote attestation for IoT networks, not AI system safety.
FreeSBee: A Constant-Time Compilation Pipeline The authors propose FreeSBee, a compiler pipeline combining taint analysis and automated control-flow transformations to defend against timing side-channels. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Deals with compiler hardening against timing attacks, not AI system safety.
Secret Sharing from Non-Extractable Randomness This paper proves that two-out-of-two secret sharing is achievable from non-extractable randomness sources, answering an open cryptographic question. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Pure cryptographic theory on secret sharing, unrelated to AI safety.
Towards Efficient Threat Intelligence Instruction Tuning via Quality-Aware and Optimal Transport-Guided Data Selection The authors present CTI-TDS, a data selection method using quality filtering and optimal transport to improve instruction tuning for threat intelligence. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on instruction tuning for cyber threat intelligence, not AI safety.
Building Sound Evaluation Practices for Rust-Based Security Frameworks This paper examines evaluation practices for Rust security testing tools and proposes a methodology to support scientifically sound assessments. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on Rust security testing frameworks, not AI model deployment safety.
Phantomware: a Data Augmentation Framework for the Detection of Emerging Malware Campaigns The authors propose Phantomware, a data augmentation framework designed to generate realistic malware variants for detecting emerging malicious campaigns. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Covers malware detection data augmentation, not AI deployment safety or LLMs.
Tulva: Practical Asynchronous Threshold FHE This paper presents Tulva, a modular asynchronous threshold fully homomorphic encryption scheme evaluated for privacy-preserving smart contract applications. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Deals with threshold fully homomorphic encryption, not AI safety or models.
Quantum-Aware Key Rank for Side-Channel Attacks evaluation The authors introduce a cost model and metric for evaluating side-channel key rank against an adversary equipped with a quantum computer. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on cryptographic key side-channel attacks, not AI safety or LLMs.
Prioritizing Microservice Requests via Retries This paper introduces DiffTry, a stateless retry mechanism that transforms retries into request prioritization to prevent microservice overload storms. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Presents cloud microservice networking retry policies, not AI deployment safety.
Can LLMs Access their World Knowledge for Event Prediction? The authors introduce DescribeEvent to show that language models can forecast temporal events when using natural-language descriptions and symbolic labels. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Does not explore small models, pruning, distillation, or efficient inference.
Can LLMs Access their World Knowledge for Event Prediction? The authors introduce DescribeEvent to show that language models can forecast temporal events when using natural-language descriptions and symbolic labels. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Studies knowledge elicitation interfaces, not safety, hallucination, or guardrails.
Can LLMs Access their World Knowledge for Event Prediction? The authors introduce DescribeEvent to show that language models can forecast temporal events when using natural-language descriptions and symbolic labels. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Evaluates natural language event forecasting, not multimodal vision-language reasoning.
Can LLMs Access their World Knowledge for Event Prediction? The authors introduce DescribeEvent to show that language models can forecast temporal events when using natural-language descriptions and symbolic labels. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Investigates prompt interface design for forecasting, not multi-step agents.
Can LLMs Access their World Knowledge for Event Prediction? The authors introduce DescribeEvent to show that language models can forecast temporal events when using natural-language descriptions and symbolic labels. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Examines temporal event forecasting, not software programming or code tasks.
Can LLMs Access their World Knowledge for Event Prediction? The authors introduce DescribeEvent to show that language models can forecast temporal events when using natural-language descriptions and symbolic labels. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Focuses on formatting event descriptions, not retrieval-augmented generation systems.
Joint Subspace Sparse Attention for Efficient Long-Context Decoding The paper introduces a token-level sparse attention method using a learned query-key subspace and an indexed kernel to accelerate long-context decoding. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Proposes an attention mechanism, not a general benchmark or evaluation methodology.
Joint Subspace Sparse Attention for Efficient Long-Context Decoding The paper introduces a token-level sparse attention method using a learned query-key subspace and an indexed kernel to accelerate long-context decoding. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Addresses attention computation efficiency, not autonomous agentic workflows.
Joint Subspace Sparse Attention for Efficient Long-Context Decoding The paper introduces a token-level sparse attention method using a learned query-key subspace and an indexed kernel to accelerate long-context decoding. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Focuses on text long-context decoding, not multimodal vision-language tasks.
Joint Subspace Sparse Attention for Efficient Long-Context Decoding The paper introduces a token-level sparse attention method using a learned query-key subspace and an indexed kernel to accelerate long-context decoding. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Focuses on attention kernel efficiency, not software engineering or coding.
Joint Subspace Sparse Attention for Efficient Long-Context Decoding The paper introduces a token-level sparse attention method using a learned query-key subspace and an indexed kernel to accelerate long-context decoding. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Addresses inference acceleration, not safety, alignment, or guardrail deployment.
Joint Subspace Sparse Attention for Efficient Long-Context Decoding The paper introduces a token-level sparse attention method using a learned query-key subspace and an indexed kernel to accelerate long-context decoding. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Optimizes attention mechanisms, not external knowledge retrieval or RAG.
Crossing the Gender Binary: Modelling Gendered Reference in Science Fiction and General Fiction The paper analyzes gendered reference replacements across science fiction and general fiction corpora using a masked language model. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Studies literary fiction corpora, not software engineering or code.
Crossing the Gender Binary: Modelling Gendered Reference in Science Fiction and General Fiction The paper analyzes gendered reference replacements across science fiction and general fiction corpora using a masked language model. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Investigates gendered reference modeling, not planning or autonomous agents.
Crossing the Gender Binary: Modelling Gendered Reference in Science Fiction and General Fiction The paper analyzes gendered reference replacements across science fiction and general fiction corpora using a masked language model. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Does not address retrieval-augmented generation or grounding mechanisms.
Crossing the Gender Binary: Modelling Gendered Reference in Science Fiction and General Fiction The paper analyzes gendered reference replacements across science fiction and general fiction corpora using a masked language model. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Focuses on computational literary analysis rather than general LLM evaluation benchmarks.
Crossing the Gender Binary: Modelling Gendered Reference in Science Fiction and General Fiction The paper analyzes gendered reference replacements across science fiction and general fiction corpora using a masked language model. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Analyzes written fiction text corpora, not multimodal vision-language data.
Crossing the Gender Binary: Modelling Gendered Reference in Science Fiction and General Fiction The paper analyzes gendered reference replacements across science fiction and general fiction corpora using a masked language model. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Analyzes literary gender bias, not AI safety or deployment guardrails.
Crossing the Gender Binary: Modelling Gendered Reference in Science Fiction and General Fiction The paper analyzes gendered reference replacements across science fiction and general fiction corpora using a masked language model. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Uses RoBERTa for masking without proposing model efficiency methods.
Tracing Morphology, Semantics, and Divergence in Language Acquisition The paper analyzes intermediate training checkpoints to trace when language models acquire morphological structures versus semantic embeddings during pretraining. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Studies language acquisition dynamics during training, not evaluation benchmarks or protocols.
Tracing Morphology, Semantics, and Divergence in Language Acquisition The paper analyzes intermediate training checkpoints to trace when language models acquire morphological structures versus semantic embeddings during pretraining. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Investigates linguistic representation learning, not multi-step autonomous agents.
Tracing Morphology, Semantics, and Divergence in Language Acquisition The paper analyzes intermediate training checkpoints to trace when language models acquire morphological structures versus semantic embeddings during pretraining. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Analyzes linguistic acquisition in LLMs, not programming or code synthesis.
Tracing Morphology, Semantics, and Divergence in Language Acquisition The paper analyzes intermediate training checkpoints to trace when language models acquire morphological structures versus semantic embeddings during pretraining. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Studies checkpoints across scales without proposing efficiency or compression techniques.
Tracing Morphology, Semantics, and Divergence in Language Acquisition The paper analyzes intermediate training checkpoints to trace when language models acquire morphological structures versus semantic embeddings during pretraining. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Investigates linguistic acquisition dynamics, not safety, jailbreaks, or deployment risks.
Tracing Morphology, Semantics, and Divergence in Language Acquisition The paper analyzes intermediate training checkpoints to trace when language models acquire morphological structures versus semantic embeddings during pretraining. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Focuses exclusively on text-based language acquisition, not multimodal models.
Tracing Morphology, Semantics, and Divergence in Language Acquisition The paper analyzes intermediate training checkpoints to trace when language models acquire morphological structures versus semantic embeddings during pretraining. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Does not involve external document retrieval or grounding architectures.
Pre-Trained Language Models as Zero-Shot Tools for Social Science Research The paper examines using pre-trained language models without task-specific training for social science record linkage and demographic inference from names. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 15:06	Not relevant Applies pre-trained models to empirical social science tasks rather than benchmarking LLMs.
Pre-Trained Language Models as Zero-Shot Tools for Social Science Research The paper examines using pre-trained language models without task-specific training for social science record linkage and demographic inference from names. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Evaluates zero-shot social science tasks, not multi-step agentic systems.
Pre-Trained Language Models as Zero-Shot Tools for Social Science Research The paper examines using pre-trained language models without task-specific training for social science record linkage and demographic inference from names. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on demographic inference accuracy, not deployment safety or guardrails.
Pre-Trained Language Models as Zero-Shot Tools for Social Science Research The paper examines using pre-trained language models without task-specific training for social science record linkage and demographic inference from names. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Introduces record linkage matching, not retrieval-augmented generation pipelines.
Pre-Trained Language Models as Zero-Shot Tools for Social Science Research The paper examines using pre-trained language models without task-specific training for social science record linkage and demographic inference from names. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Applies models to record linkage and demographic inference, not programming.
Pre-Trained Language Models as Zero-Shot Tools for Social Science Research The paper examines using pre-trained language models without task-specific training for social science record linkage and demographic inference from names. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Operates exclusively on tabular and text data, not multimodal inputs.
Pre-Trained Language Models as Zero-Shot Tools for Social Science Research The paper examines using pre-trained language models without task-specific training for social science record linkage and demographic inference from names. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Mentions open-weight models in passing without focusing on efficiency techniques.
Similar Models, Broken Odds: Where Luce's Axiom Fails in Chatbot Arena The paper shows that the Bradley-Terry model fails on Chatbot Arena for similar models, and introduces a parameter-efficient correction for ties. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Focuses on general chatbot evaluation, not software engineering or code.
Similar Models, Broken Odds: Where Luce's Axiom Fails in Chatbot Arena The paper shows that the Bradley-Terry model fails on Chatbot Arena for similar models, and introduces a parameter-efficient correction for ties. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Examines benchmark methodology rather than deployment safety, jailbreaks, or hallucinations.
Similar Models, Broken Odds: Where Luce's Axiom Fails in Chatbot Arena The paper shows that the Bradley-Terry model fails on Chatbot Arena for similar models, and introduces a parameter-efficient correction for ties. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Does not investigate model compression, distillation, or efficient serving.
Similar Models, Broken Odds: Where Luce's Axiom Fails in Chatbot Arena The paper shows that the Bradley-Terry model fails on Chatbot Arena for similar models, and introduces a parameter-efficient correction for ties. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Studies ranking models for arena votes, not autonomous agent workflows.
Similar Models, Broken Odds: Where Luce's Axiom Fails in Chatbot Arena The paper shows that the Bradley-Terry model fails on Chatbot Arena for similar models, and introduces a parameter-efficient correction for ties. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Does not touch on retrieval augmentation or external knowledge retrieval.
Similar Models, Broken Odds: Where Luce's Axiom Fails in Chatbot Arena The paper shows that the Bradley-Terry model fails on Chatbot Arena for similar models, and introduces a parameter-efficient correction for ties. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Examines Chatbot Arena text rankings, not multimodal model architectures.
DecisionRobust: Benchmarking Robustness for JEV Style Decision Models Across Text and Vision The paper introduces a benchmark evaluating the robustness of typed decision models against text, image, and prompt injection attacks. Preprint	Small and efficient models via arxiv · 15 Oct, 15:06	Not relevant Does not address model efficiency, distillation, or quantization techniques.
DecisionRobust: Benchmarking Robustness for JEV Style Decision Models Across Text and Vision The paper introduces a benchmark evaluating the robustness of typed decision models against text, image, and prompt injection attacks. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 15:06	Not relevant Does not cover retrieval mechanisms or external database grounding.
DecisionRobust: Benchmarking Robustness for JEV Style Decision Models Across Text and Vision The paper introduces a benchmark evaluating the robustness of typed decision models against text, image, and prompt injection attacks. Preprint	LLMs for code via arxiv · 15 Oct, 15:06	Not relevant Evaluates decision models on text and vision, not software code generation.
DecisionRobust: Benchmarking Robustness for JEV Style Decision Models Across Text and Vision The paper introduces a benchmark evaluating the robustness of typed decision models against text, image, and prompt injection attacks. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Focuses on decision model robustness, not autonomous planning agents.
DivAlgo: A systematic evaluation of algorithmic diversity for code generation models The paper introduces a benchmark and similarity metric to measure the algorithmic diversity of solutions produced by code generation models. Preprint	Multimodal models via arxiv · 15 Oct, 15:06	Not relevant Examines source code and text, not multimodal image or video understanding.
DivAlgo: A systematic evaluation of algorithmic diversity for code generation models The paper introduces a benchmark and similarity metric to measure the algorithmic diversity of solutions produced by code generation models. Preprint	AI agents via arxiv · 15 Oct, 15:06	Not relevant Evaluates code diversity without investigating multi-step planning or tool use.
DivAlgo: A systematic evaluation of algorithmic diversity for code generation models The paper introduces a benchmark and similarity metric to measure the algorithmic diversity of solutions produced by code generation models. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 15:06	Not relevant Focuses on algorithmic diversity in code, not safety or guardrails.

6,601–6,660 of 17,313← Previous Next →