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
Robust Marathi Sentiment Analysis using Domain-Adaptive Pre-training and Data Augmentation Approaches The authors evaluate domain-adaptive pretraining and back-translation data augmentation to improve sentiment classification in Marathi. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:18	Not relevant Does not evaluate safety, jailbreaks, or deployment risks.
Robust Marathi Sentiment Analysis using Domain-Adaptive Pre-training and Data Augmentation Approaches The authors evaluate domain-adaptive pretraining and back-translation data augmentation to improve sentiment classification in Marathi. Preprint	AI agents via arxiv · 15 Oct, 06:18	Not relevant Focuses on text classification, not AI agents.
Robust Marathi Sentiment Analysis using Domain-Adaptive Pre-training and Data Augmentation Approaches The authors evaluate domain-adaptive pretraining and back-translation data augmentation to improve sentiment classification in Marathi. Preprint	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Deals with Marathi text sentiment, not code.
Robust Marathi Sentiment Analysis using Domain-Adaptive Pre-training and Data Augmentation Approaches The authors evaluate domain-adaptive pretraining and back-translation data augmentation to improve sentiment classification in Marathi. Preprint	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Not focused on model compression or efficient inference.
Robust Marathi Sentiment Analysis using Domain-Adaptive Pre-training and Data Augmentation Approaches The authors evaluate domain-adaptive pretraining and back-translation data augmentation to improve sentiment classification in Marathi. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Text-only classification study.
RSIPO: Recursive Self-Improving Policy Optimization for Agentic Reinforcement Learning via Skill-Policy Co-Evolution RSIPO introduces a reinforcement learning framework where agentic skills and policy models co-evolve using self-extracted trajectory experience. Preprint	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Not focused on code synthesis or software engineering.
RSIPO: Recursive Self-Improving Policy Optimization for Agentic Reinforcement Learning via Skill-Policy Co-Evolution RSIPO introduces a reinforcement learning framework where agentic skills and policy models co-evolve using self-extracted trajectory experience. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:18	Not relevant Addresses reinforcement learning policies rather than RAG systems.
RSIPO: Recursive Self-Improving Policy Optimization for Agentic Reinforcement Learning via Skill-Policy Co-Evolution RSIPO introduces a reinforcement learning framework where agentic skills and policy models co-evolve using self-extracted trajectory experience. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:18	Not relevant Does not introduce a new benchmark or judge method.
RSIPO: Recursive Self-Improving Policy Optimization for Agentic Reinforcement Learning via Skill-Policy Co-Evolution RSIPO introduces a reinforcement learning framework where agentic skills and policy models co-evolve using self-extracted trajectory experience. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Focuses on policy optimization, not multimodal models.
RSIPO: Recursive Self-Improving Policy Optimization for Agentic Reinforcement Learning via Skill-Policy Co-Evolution RSIPO introduces a reinforcement learning framework where agentic skills and policy models co-evolve using self-extracted trajectory experience. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:18	Not relevant Focuses on task performance rather than safety guardrails.
RSIPO: Recursive Self-Improving Policy Optimization for Agentic Reinforcement Learning via Skill-Policy Co-Evolution RSIPO introduces a reinforcement learning framework where agentic skills and policy models co-evolve using self-extracted trajectory experience. Preprint	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Not focused on model quantization, pruning, or efficiency.
Who Sets the Pace of Judge Hacking? Judge Capability versus the Policy-Judge Gap This work analyzes whether policy capability gaps or judge models dictate reward hacking when training models against language-model judges. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Exclusively considers text LLMs and text judges.
Who Sets the Pace of Judge Hacking? Judge Capability versus the Policy-Judge Gap This work analyzes whether policy capability gaps or judge models dictate reward hacking when training models against language-model judges. Preprint	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Focuses on judge capability and RL, not compression.
Who Sets the Pace of Judge Hacking? Judge Capability versus the Policy-Judge Gap This work analyzes whether policy capability gaps or judge models dictate reward hacking when training models against language-model judges. Preprint	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Not focused on code generation or software engineering.
Who Sets the Pace of Judge Hacking? Judge Capability versus the Policy-Judge Gap This work analyzes whether policy capability gaps or judge models dictate reward hacking when training models against language-model judges. Preprint	AI agents via arxiv · 15 Oct, 06:18	Not relevant Studies RL judge exploitation, not autonomous tool agents.
Who Sets the Pace of Judge Hacking? Judge Capability versus the Policy-Judge Gap This work analyzes whether policy capability gaps or judge models dictate reward hacking when training models against language-model judges. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:18	Not relevant Does not address retrieval or external database augmentation.
Heterogeneous Structured Pruning Under Resource Constraints for Hardware-Efficient Large Language Models Heterogeneous Structured Pruning selects sublayer-specific sparsity patterns guided by latency and memory models for hardware-efficient inference. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:18	Not relevant Focuses on model compression, not retrieval or grounding.
Heterogeneous Structured Pruning Under Resource Constraints for Hardware-Efficient Large Language Models Heterogeneous Structured Pruning selects sublayer-specific sparsity patterns guided by latency and memory models for hardware-efficient inference. Preprint	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Not focused on code assistants or programming models.
Heterogeneous Structured Pruning Under Resource Constraints for Hardware-Efficient Large Language Models Heterogeneous Structured Pruning selects sublayer-specific sparsity patterns guided by latency and memory models for hardware-efficient inference. Preprint	AI agents via arxiv · 15 Oct, 06:18	Not relevant Hardware pruning method, not related to AI agents.
Heterogeneous Structured Pruning Under Resource Constraints for Hardware-Efficient Large Language Models Heterogeneous Structured Pruning selects sublayer-specific sparsity patterns guided by latency and memory models for hardware-efficient inference. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:18	Not relevant Does not introduce LLM benchmarks or evaluation methods.
Heterogeneous Structured Pruning Under Resource Constraints for Hardware-Efficient Large Language Models Heterogeneous Structured Pruning selects sublayer-specific sparsity patterns guided by latency and memory models for hardware-efficient inference. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:18	Not relevant Focuses on latency and memory efficiency, not safety.
Heterogeneous Structured Pruning Under Resource Constraints for Hardware-Efficient Large Language Models Heterogeneous Structured Pruning selects sublayer-specific sparsity patterns guided by latency and memory models for hardware-efficient inference. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Evaluates text LLMs, not multimodal architectures.
Local Citation Recommendation Without Graph Signals The paper introduces a dataset and citation recommendation system that retrieves relevant scientific papers without relying on citation graphs. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:18	Not relevant Focuses on retrieval accuracy, not AI safety risks.
Local Citation Recommendation Without Graph Signals The paper introduces a dataset and citation recommendation system that retrieves relevant scientific papers without relying on citation graphs. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Processes textual papers only, no multimodal content.
Local Citation Recommendation Without Graph Signals The paper introduces a dataset and citation recommendation system that retrieves relevant scientific papers without relying on citation graphs. Preprint	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Does not study model pruning, quantization, or distillation.
Local Citation Recommendation Without Graph Signals The paper introduces a dataset and citation recommendation system that retrieves relevant scientific papers without relying on citation graphs. Preprint	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Focuses on scientific literature, not code.
Local Citation Recommendation Without Graph Signals The paper introduces a dataset and citation recommendation system that retrieves relevant scientific papers without relying on citation graphs. Preprint	AI agents via arxiv · 15 Oct, 06:18	Not relevant Presents citation retrieval, not autonomous planning agents.
Latent Taxonomies: How Property Induction Develops in Language Model Pre-training This study analyzes how category-based property induction emerges in language models during pretraining using a synthetic hierarchical taxonomy. Preprint	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Not focused on code generation or software engineering.
Latent Taxonomies: How Property Induction Develops in Language Model Pre-training This study analyzes how category-based property induction emerges in language models during pretraining using a synthetic hierarchical taxonomy. Preprint	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Analyzes scaling effects, not compression or efficient inference.
Latent Taxonomies: How Property Induction Develops in Language Model Pre-training This study analyzes how category-based property induction emerges in language models during pretraining using a synthetic hierarchical taxonomy. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:18	Not relevant Studies pre-training phenomena rather than testing benchmarks.
Latent Taxonomies: How Property Induction Develops in Language Model Pre-training This study analyzes how category-based property induction emerges in language models during pretraining using a synthetic hierarchical taxonomy. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:18	Not relevant Not focused on safety, jailbreaks, or deployment guardrails.
Latent Taxonomies: How Property Induction Develops in Language Model Pre-training This study analyzes how category-based property induction emerges in language models during pretraining using a synthetic hierarchical taxonomy. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:18	Not relevant Focuses on pretraining memory and induction, not RAG.
Latent Taxonomies: How Property Induction Develops in Language Model Pre-training This study analyzes how category-based property induction emerges in language models during pretraining using a synthetic hierarchical taxonomy. Preprint	AI agents via arxiv · 15 Oct, 06:18	Not relevant Studies pretraining dynamics, not multi-step agent behavior.
Latent Taxonomies: How Property Induction Develops in Language Model Pre-training This study analyzes how category-based property induction emerges in language models during pretraining using a synthetic hierarchical taxonomy. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Investigates textual taxonomies only, not multimodal data.
KhyontekAI at CheckThat! 2026: Multilingual Dense Retrieval and Cross-Encoder Reranking for Scientific Web Claims The authors present a multilingual dense retrieval and reranking framework to link scientific web claims to their source publications. Peer-reviewed	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Focuses on scientific claims, not software or code.
KhyontekAI at CheckThat! 2026: Multilingual Dense Retrieval and Cross-Encoder Reranking for Scientific Web Claims The authors present a multilingual dense retrieval and reranking framework to link scientific web claims to their source publications. Peer-reviewed	AI safety and reliability in deployment via arxiv · 15 Oct, 06:18	Not relevant Does not evaluate safety, jailbreaks, or guardrails.
KhyontekAI at CheckThat! 2026: Multilingual Dense Retrieval and Cross-Encoder Reranking for Scientific Web Claims The authors present a multilingual dense retrieval and reranking framework to link scientific web claims to their source publications. Peer-reviewed	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:18	Not relevant Presents a competition system, not an evaluation framework.
KhyontekAI at CheckThat! 2026: Multilingual Dense Retrieval and Cross-Encoder Reranking for Scientific Web Claims The authors present a multilingual dense retrieval and reranking framework to link scientific web claims to their source publications. Peer-reviewed	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Deals exclusively with text and metadata.
KhyontekAI at CheckThat! 2026: Multilingual Dense Retrieval and Cross-Encoder Reranking for Scientific Web Claims The authors present a multilingual dense retrieval and reranking framework to link scientific web claims to their source publications. Peer-reviewed	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Does not address model compression or efficient serving.
KhyontekAI at CheckThat! 2026: Multilingual Dense Retrieval and Cross-Encoder Reranking for Scientific Web Claims The authors present a multilingual dense retrieval and reranking framework to link scientific web claims to their source publications. Peer-reviewed	AI agents via arxiv · 15 Oct, 06:18	Not relevant Document retrieval framework, not an autonomous agent system.
Binding Phoneme Identity to Position in Neural Models and the Human Brain The study tests whether neural models bind phoneme identity to serial position using state direction and magnitude representations. Preprint	AI agents via arxiv · 15 Oct, 06:18	Not relevant Cognitive and representational neuroscience paper, not about agents.
Binding Phoneme Identity to Position in Neural Models and the Human Brain The study tests whether neural models bind phoneme identity to serial position using state direction and magnitude representations. Preprint	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Not focused on programming or code synthesis.
Binding Phoneme Identity to Position in Neural Models and the Human Brain The study tests whether neural models bind phoneme identity to serial position using state direction and magnitude representations. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:18	Not relevant Not related to retrieval or document augmentation.
Binding Phoneme Identity to Position in Neural Models and the Human Brain The study tests whether neural models bind phoneme identity to serial position using state direction and magnitude representations. Preprint	LLM evaluation and benchmarks via arxiv · 15 Oct, 06:18	Not relevant Does not introduce benchmarks or LLM evaluation frameworks.
Binding Phoneme Identity to Position in Neural Models and the Human Brain The study tests whether neural models bind phoneme identity to serial position using state direction and magnitude representations. Preprint	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Investigates phoneme representations, not model efficiency.
Binding Phoneme Identity to Position in Neural Models and the Human Brain The study tests whether neural models bind phoneme identity to serial position using state direction and magnitude representations. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Focuses on text and phonemes, not multimodal models.
Binding Phoneme Identity to Position in Neural Models and the Human Brain The study tests whether neural models bind phoneme identity to serial position using state direction and magnitude representations. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:18	Not relevant Does not address deployment safety, hallucination, or alignment.
What Do Sparse Autoencoder Feature Discovery Scores Identify? From Activations to Interventions This paper compares observational feature discovery scores in sparse autoencoders against actual causal intervention effects on model outputs. Preprint	AI agents via arxiv · 15 Oct, 06:18	Not relevant Mechanistic interpretability study, not about agentic workflows.
What Do Sparse Autoencoder Feature Discovery Scores Identify? From Activations to Interventions This paper compares observational feature discovery scores in sparse autoencoders against actual causal intervention effects on model outputs. Preprint	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Not about coding assistants or software engineering tasks.
What Do Sparse Autoencoder Feature Discovery Scores Identify? From Activations to Interventions This paper compares observational feature discovery scores in sparse autoencoders against actual causal intervention effects on model outputs. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:18	Not relevant Not related to retrieval-augmented generation.
What Do Sparse Autoencoder Feature Discovery Scores Identify? From Activations to Interventions This paper compares observational feature discovery scores in sparse autoencoders against actual causal intervention effects on model outputs. Preprint	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Studies feature steering and ablation, not model compression.
What Do Sparse Autoencoder Feature Discovery Scores Identify? From Activations to Interventions This paper compares observational feature discovery scores in sparse autoencoders against actual causal intervention effects on model outputs. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Does not involve multimodal representations or models.
Unchecked or Falsely Accused? Evaluating AI-Generated Text Detectors on Kazakh This work evaluates machine-text detectors on Kazakh text using pre-ChatGPT human writing and shows significant false-positive rates. Preprint	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Focuses on natural language text, not code generation.
Unchecked or Falsely Accused? Evaluating AI-Generated Text Detectors on Kazakh This work evaluates machine-text detectors on Kazakh text using pre-ChatGPT human writing and shows significant false-positive rates. Preprint	Multimodal models via arxiv · 15 Oct, 06:18	Not relevant Text-only dataset and detectors, no multimodal components.
Unchecked or Falsely Accused? Evaluating AI-Generated Text Detectors on Kazakh This work evaluates machine-text detectors on Kazakh text using pre-ChatGPT human writing and shows significant false-positive rates. Preprint	AI agents via arxiv · 15 Oct, 06:18	Not relevant Evaluates text detectors, not AI agents.
Unchecked or Falsely Accused? Evaluating AI-Generated Text Detectors on Kazakh This work evaluates machine-text detectors on Kazakh text using pre-ChatGPT human writing and shows significant false-positive rates. Preprint	Retrieval-augmented generation (RAG) via arxiv · 15 Oct, 06:18	Not relevant Does not address retrieval-augmented generation.
Unchecked or Falsely Accused? Evaluating AI-Generated Text Detectors on Kazakh This work evaluates machine-text detectors on Kazakh text using pre-ChatGPT human writing and shows significant false-positive rates. Preprint	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Not related to model efficiency, quantization, or distillation.
RIVO: Lightweight Adaptive Rerouting for Deep Research Agents RIVO is a lightweight plug-in framework that adaptively reroutes research agent trajectories to prevent premature stopping or wasted steps. Preprint	LLMs for code via arxiv · 15 Oct, 06:18	Not relevant Not focused on code generation or software engineering tasks.
RIVO: Lightweight Adaptive Rerouting for Deep Research Agents RIVO is a lightweight plug-in framework that adaptively reroutes research agent trajectories to prevent premature stopping or wasted steps. Preprint	Small and efficient models via arxiv · 15 Oct, 06:18	Not relevant Does not study model compression, quantization, or efficient architectures.
RIVO: Lightweight Adaptive Rerouting for Deep Research Agents RIVO is a lightweight plug-in framework that adaptively reroutes research agent trajectories to prevent premature stopping or wasted steps. Preprint	AI safety and reliability in deployment via arxiv · 15 Oct, 06:18	Not relevant Focuses on task completion utility rather than safety or reliability.

10,801–10,860 of 17,313← Previous Next →