#research
Every summary, chronological. Filter by category, tag, or source from the rail.
The Accuracy-Efficiency Paradox in On-Device Energy Forecasting
On-device energy forecasting models often consume more power than the energy savings they aim to provide, creating a net-negative efficiency paradox that requires careful calibration of model complexity.
Standardizing Distributed AI Workflows with SAREF Ontologies
The article proposes an ontology based on the Smart Applications REFerence (SAREF) standard to enable interoperability and orchestration of AI workflows across edge, fog, and cloud computing environments.
EEG-to-Report: Bridging Clinical Brain Data and Language Models
The EEG-to-Report framework introduces a standardized annotation and feature-text mapping method to enable training language models on complex clinical EEG data, bridging the gap between raw neural signals and diagnostic reports.
Building Safe Multimodal AI for Mental Health Support
The Anian framework introduces a safety-gated architecture for mental health AI, utilizing hierarchical state representation and conservative risk fusion to ensure controlled, reliable patient interactions.
Knowledge Cards: A Framework for Structured AI Knowledge
Knowledge Cards provide a standardized, machine-readable format for documenting AI model capabilities, limitations, and provenance, moving beyond unstructured documentation to improve transparency and reliability.
Refusal Is Not Robustness: LLMs Fabricate on Uninformative Data
Large Language Models often fail to identify uninformative input, choosing to confidently fabricate clinical assessments rather than admitting a lack of sufficient data.
The 5D Framework for Multi-Table Data Analysis
The 5D framework provides a unified methodology for integrating and reusing complex, multi-table datasets by mapping data across five distinct dimensions to ensure consistency and analytical depth.
Explaining ICU Mortality Predictions with LLM Agentic Pipelines
This study demonstrates the feasibility of using standalone LLMs and pre-specified agentic pipelines to interpret complex ICU mortality risk models, providing a path toward more transparent clinical decision support.
EduRiskX: Combining Transformers and F-Logic for Academic Prediction
EduRiskX improves academic risk prediction by pairing temporal Transformers for pattern recognition with F-Logic for rule-based, interpretable reasoning.
Anthropic's Automated Researcher: A Leap in Self-Improving AI
Anthropic researchers have developed an Automated Alignment Researcher (AAR) that outperforms human researchers at improving model alignment, doing so at a fraction of the cost and time.
Formal Verification for AI-Generated Code with Lean4
As AI agents generate code at scale, traditional testing and human review fail to guarantee correctness. Formal verification using Lean4 allows developers to define specifications that machines prove mathematically, ensuring code is correct for every possible input.
A Formal Framework for Auditing XAI Robustness and Fidelity
This paper proposes a formal methodology to audit Explainable AI (XAI) systems, ensuring that explanations are both robust to input perturbations and faithful to the underlying model's decision-making process.
Decoupling Model Performance from Evaluation Bias
Current AI benchmarks often conflate model capability with the biases of the evaluation instrument itself, necessitating a shift toward disentangling model preferences from measurement artifacts.
Optimizing Code Models with Function-Level Execution Feedback
Improving code generation models by using granular, function-level execution feedback rather than binary pass/fail signals to guide preference optimization.
RL-Enhanced Agentic Search for Biomedical Fact-Checking
This paper introduces a reinforcement learning-based agentic framework designed to improve the accuracy and reliability of automated biomedical fact-checking by optimizing search strategies.
Reducing Medical AI Sycophancy via Gated Activation Steering
Gated Activation Steering (GAS) improves medical LLM reliability by dynamically suppressing internal representations associated with sycophancy and hallucinations during inference, without requiring model retraining.
Optimizing Masked Diffusion LLMs for Real-World Hardware
This paper provides a characterization of Masked Diffusion LLMs, identifying unique computational bottlenecks and proposing hardware-aware design principles to improve inference efficiency.
RENDER: A Framework for Controlling Evidence in LLM Memory Evaluation
RENDER is a new evaluation framework designed to isolate and measure how LLMs process and recall specific evidence within their context windows, addressing the limitations of existing memory benchmarks.
Evaluating NL2SQL Performance with ESQ-Bench
ESQ-Bench is a new benchmark designed to test NL2SQL models on dialect generalization and silent semantic divergence, addressing the limitations of existing benchmarks in enterprise environments.
Architecture-Aware Credit Transport for LLM Reinforcement Learning
The paper introduces a method to improve LLM reinforcement learning by aligning credit assignment with the underlying computational architecture, ensuring rewards are distributed based on actual processing paths.
Composable Trust Infrastructure for Manufacturing Knowledge Graphs
This paper proposes a framework for integrating cross-system provenance, temporal reasoning, and decision traceability into manufacturing knowledge graphs to ensure reliable AI-driven industrial operations.
Agentic AI in Safety-Critical Multi-Drone Systems
Integrating agentic AI into multi-drone systems requires balancing autonomous decision-making with strict safety constraints, human-in-the-loop oversight, and robust verification methods.
LitReview Arena: Benchmarking AI Agents for Literature Synthesis
LitReview Arena introduces a battle-style evaluation platform to measure the accuracy, synthesis capabilities, and citation integrity of AI agents performing academic literature reviews.
Hate Speech Classification in Roman Urdu: PEFT vs. Prompt Engineering
A comparative study evaluating Parameter-Efficient Fine-Tuning (PEFT) against prompt engineering for detecting hate speech in Roman Urdu, highlighting the trade-offs between computational efficiency and classification accuracy in low-resource linguistic contexts.
Frameworks for Explainable AI in Time Series Classification
A systematic review of current software frameworks for XAI in time series classification, highlighting the need for standardized evaluation and better integration of interpretability tools in production pipelines.
AIREP: A Protocol for Verifiable AI Runtime Governance
AIREP (AI Runtime Evidence Protocol) provides a standardized framework for generating and verifying cryptographic evidence for individual AI decisions, enabling transparent and auditable runtime governance.
Designing AI Environments for Collective Intelligence
Moving from rigid agent workflows to open, incentive-driven environments enables AI to solve complex scientific problems and optimize GPU kernels through collective, iterative collaboration.
AI EngineerStructurally Indirect Prerequisite Eviction in Agentic Memory
Agentic memory systems often fail not due to retrieval errors, but because 'prerequisite' information is evicted from context before it can be used, creating a structural failure in long-term reasoning.
Consilience: Improving Multi-Agent Reasoning via Calibration
Consilience introduces a framework for multi-agent systems to solve hidden-profile problems by using conformal calibration to control communication and reduce information bias.
Defining World Models for Agents and Environments
This paper provides a formal framework for understanding world models by distinguishing between environment-only, agent-only, and joint agent-environment system dynamics.
Showing 30 of 375