3/12/2026
Over-Searching in Search-Augmented Large Language Models
Evaluates over-searching behavior in retrieval-augmented LLMs where models invoke search tools unnecessarily, causing hallucinations and inefficiency.
3/12/2026
Time series forecasting with Hahn Kolmogorov-Arnold networks
Introduces HaKAN, a Kolmogorov-Arnold Network model for time series forecasting with learnable activation functions and lower computational complexity.
3/12/2026
Hallucination is a Consequence of Space-Optimality: A Rate-Distortion Theorem for Membership Testing
Theoretical analysis of LLM hallucinations using rate-distortion theory and membership testing, explaining confidence on unfactual claims.
3/12/2026
Position: Beyond Model-Centric Prediction -- Agentic Time Series Forecasting
Proposes agentic approach to time series forecasting using iterative refinement and reasoning instead of static model-centric predictions.
3/12/2026
Grounding Generated Videos in Feasible Plans via World Models
GVP-WM grounds video-generated plans using world models to ensure temporal consistency and physical feasibility for robotic action execution.
3/12/2026
Expert-Data Alignment Governs Generation Quality in Decentralized Diffusion Models
Analysis of decentralized diffusion models showing how expert-data alignment governs generation quality when experts trained on disjoint data clusters.
3/12/2026
BLITZRANK: Principled Zero-shot Ranking Agents with Tournament Graphs
Zero-shot ranking algorithm using tournament graphs for k-wise comparisons, applicable to LLM-based document reranking and evaluation.
3/12/2026
Latent Poincar\'e Shaping for Agentic Reinforcement Learning
Method for training AlphaZero-like LLM agents in hyperbolic Poincaré space, using exponentially increasing latent capacity for tree search.
3/12/2026
Long Chain-of-Thought Compression via Fine-Grained Group Policy Optimization
RL algorithm that compresses verbose chain-of-thought reasoning in LLMs through fine-grained group policy optimization based on length and entropy.
3/12/2026
LexiSafe: Offline Safe Reinforcement Learning with Lexicographic Safety-Reward Hierarchy
Offline safe reinforcement learning framework using lexicographic safety-reward hierarchy to prevent safety drift in cyber-physical systems.
3/12/2026
Solving adversarial examples requires solving exponential misalignment
Analysis of adversarial examples through neural network perceptual manifolds, showing dimensionality misalignment between manifolds causes vulnerability.
3/12/2026
BD-Merging: Bias-Aware Dynamic Model Merging with Evidence-Guided Contrastive Learning
Model merging method for multi-task learning that handles test-time distribution shift through bias-aware dynamic merging.
3/12/2026
Stochastic Attention via Langevin Dynamics on the Modern Hopfield Energy
Interprets attention mechanism as one step of gradient descent, enabling stochastic attention via Langevin sampling controlled by temperature.
3/12/2026
SPAARS: Safer RL Policy Alignment through Abstract Exploration and Refined Exploitation of Action Space
Safe offline-to-online reinforcement learning method for robotics using CVAEs to constrain exploration within behavioral support of offline data.
3/12/2026
Mindstorms in Natural Language-Based Societies of Mind
Societies of large language models collaborating through natural language interfaces to solve problems, inspired by Minsky's society of mind concept.
3/12/2026
Losing dimensions: Geometric memorization in generative diffusion
Research on how diffusion models memorize training data, measuring latent dimensionality collapse on low-dimensional manifolds as data becomes scarce.
3/12/2026
Scalable Multi-Task Learning through Spiking Neural Networks with Adaptive Task-Switching Policy for Intelligent Autonomous Agents
Spiking neural networks with adaptive task-switching for resource-constrained multi-task reinforcement learning in autonomous agents.
3/12/2026
LLLMs: A Data-Driven Survey of Evolving Research on Limitations of Large Language Models
Data-driven survey identifying 14,648 papers on LLM limitations from 2022-2025 using LLM-based classification.
3/12/2026
Consistency-based Abductive Reasoning over Perceptual Errors of Multiple Pre-trained Models in Novel Environments
System using multiple pre-trained models and consistency-based reasoning to handle distributional shifts in novel environments.
3/12/2026
Learning What Reinforcement Learning Can't: Interleaved Online Fine-Tuning for Hardest Questions
Interleaved online fine-tuning method enabling LLMs to acquire capabilities beyond base model through reinforcement learning.
3/12/2026
Differential Privacy in Machine Learning: A Survey from Symbolic AI to LLMs
Survey of differential privacy techniques from symbolic AI through LLMs, covering definitions and applications.
3/12/2026
The Yokai Learning Environment: Tracking Beliefs Over Space and Time
Yokai: benchmark environment for zero-shot coordination in cooperative AI with belief tracking over space and time.
3/12/2026
Multi-modal Data Spectrum: Multi-modal Datasets are Multi-dimensional
Empirical study analyzing intra-modality and inter-modality dependencies in multi-modal learning benchmark datasets.
3/12/2026
RADAR: Reasoning-Ability and Difficulty-Aware Routing for Reasoning LLMs
RADAR: dynamic routing system that selects optimal reasoning LLM models based on task difficulty and cost-performance tradeoffs.
3/12/2026
KV Cache Transform Coding for Compact Storage in LLM Inference
KVTC: transform coder that compresses KV caches for efficient LLM inference storage and memory management.
3/12/2026
CostNav: A Navigation Benchmark for Real-World Economic-Cost Evaluation of Physical AI Agents
CostNav benchmark evaluating physical AI agents on real-world economic metrics for autonomous delivery systems beyond task success.
3/12/2026
Cross-embodied Co-design for Dexterous Hands
Co-design framework jointly optimizing robot hand morphology and dexterous control policies across multiple embodiments.
3/12/2026
Toward Closed-loop Molecular Discovery via Language Model, Property Alignment and Strategic Search
Closed-loop drug discovery system using language models, property alignment, and strategic search for de novo ligand design.
3/12/2026
PvP: Data-Efficient Humanoid Robot Learning with Proprioceptive-Privileged Contrastive Representations
Proprioceptive-Privileged contrastive learning framework for sample-efficient whole-body control in humanoid robots using reinforcement learning.
3/12/2026
Gradient Dynamics of Attention: How Cross-Entropy Sculpts Bayesian Manifolds
First-order analysis of how cross-entropy training reshapes transformer attention geometry for probabilistic reasoning.
3/12/2026
Benchmarking Graph Neural Networks in Solving Hard Constraint Satisfaction Problems
Benchmarking study comparing graph neural networks against classical heuristics on hard constraint satisfaction problems.
3/12/2026
Many AI Analysts, One Dataset: Navigating the Agentic Data Science Multiverse
Autonomous AI analysts built on LLMs analyzing same dataset independently, quantifying variability in research conclusions without human coordination.
3/12/2026
How Large Language Models Get Stuck: Early structure with persistent errors
Study of how LLMs develop early syntactic structures that persist as errors through training, using OPT model on BabyLM dataset.
3/12/2026
CARE: Towards Clinical Accountability in Multi-Modal Medical Reasoning with an Evidence-Grounded Agentic Framework
CARE: Evidence-grounded agentic framework using visual language models for medical reasoning with explainability and clinical accountability.
3/12/2026
RACAS: Controlling Diverse Robots With a Single Agentic System
RACAS: Framework for controlling diverse robotic platforms with a single agentic system through unified interface.
3/12/2026
A Systematic Comparison of Training Objectives for Out-of-Distribution Detection in Image Classification
Systematic comparison of training objectives (cross-entropy, prototype, triplet, AP loss) for out-of-distribution detection in image classification models.
3/12/2026
Show HN: MCP server for ICD-10 and SNOMED clinical coding
MCP server exposing API for clinical coding enabling AI assistants to convert medical text into ICD-10 and SNOMED-CT structured codes.
3/12/2026
Ask HN: Im looking for indie hackers or small teams to test AI analytics tool
Pluk enables agentic data notebooks on personal databases with plain language queries, SQL workflows, and interactive dashboard generation running locally.
3/12/2026
Can LLMs Be Computers?
Research on executing programs inside transformer models with faster inference. Limited detail in abstract.
3/12/2026
A better agent skills manager
CLI tool for managing AI agent skills from GitHub repos and local directories. Skill discovery and linking via SKILL.md manifests.
3/12/2026
Protective Dome for AI Agents – MCP Security Gateway
MCPDome: open-source security gateway for AI agents enforcing authentication, authorization, rate limiting and injection detection on MCP servers.
3/12/2026
Laminae – Multi-Agent Cognitive Pipeline
Laminae is an open-source Rust SDK adding guardrails, safety, personality, and learning layers to LLM applications for production use.
3/12/2026
Show HN: SwarmClaw – Manage a swarm of OpenClaw agents from one self-hosted UI
SwarmClaw: Control plane for managing multiple OpenClaw agent instances. Supports 14+ LLM providers, multi-agent workflows, and chat integrations.
3/12/2026
Open-source GPU virtualization and pooling for Kubernetes
TensorFusion.AI: GPU virtualization and pooling solution for Kubernetes clusters. Infrastructure for optimizing AI compute resources.
3/12/2026
Are we doing UX for AI the right way?
Essay on UX design challenges for AI products in 2026, addressing AI fatigue and defining best practices.
3/12/2026
Show HN: K9 Audit – Causal intent-execution audit trail for AI agents
K9 Audit: open-source tool capturing intent-execution audit trails for AI agents via causal five-tuples per step.
3/12/2026
Show HN: I built an open harness that excels at autonomous ML research
Helios: Open-source harness for autonomous ML research agents. Unrestricted agent execution framework for machine learning experiments.
3/12/2026
AI should help us produce better code
Guide on agentic engineering patterns for using AI agents to improve code quality rather than reduce it.
3/12/2026
SmallClaw: Local-first AI agent framework built for small models
Open source local-first AI agent framework supporting small models with optional cloud hybrid support. v1.1 release with multi-agent capabilities.
3/12/2026
Divine-OS – Persistent Identity Layer for AI Agents
Open source middleware layer adding persistent identity, auditable safety, and governance for AI agents. 542 passing tests, production-ready.