3/31/2026
SpecMoE: Spectral Mixture-of-Experts Foundation Model for Cross-Species EEG Decoding
Mixture-of-Experts foundation model for cross-species EEG signal decoding using spectral analysis.
3/31/2026
Off-Policy Learning with Limited Supply
Off-policy learning framework for contextual bandits with constrained item supply in recommendation and advertising.
3/31/2026
Wasserstein Propagation for Reverse Diffusion under Weak Log-Concavity: Exploiting Metric Mismatch via One-Switch Routing
Theoretical analysis of reverse diffusion sampling error under weak log-concavity using metric mismatch exploitation.
3/31/2026
Graph-Aware Stealthy Poison-Text Backdoors for Text-Attributed Graphs
Backdoor attack method on text-attributed graphs by injecting malicious cues into node text.
3/31/2026
SmaAT-QMix-UNet: A Parameter-Efficient Vector-Quantized UNet for Precipitation Nowcasting
Parameter-efficient vector-quantized UNet for weather precipitation nowcasting using deep learning.
3/31/2026
Decoupling Exploration and Policy Optimization: Uncertainty Guided Tree Search for Hard Exploration
RL approach decoupling exploration and policy optimization using uncertainty-guided tree search for hard exploration problems.
3/31/2026
Scaling Attention via Feature Sparsity
Sparse Feature Attention (SFA) method reducing transformer attention complexity via feature-level sparsity for ultra-long contexts.
3/31/2026
Problems with Chinchilla Approach 2: Systematic Biases in IsoFLOP Parabola Fits
Analysis of systematic biases in Chinchilla Approach 2 neural scaling law fits for compute-optimal LLM allocation.
3/31/2026
SkillRouter: Skill Routing for LLM Agents at Scale
Skill routing system for LLM agents managing and selecting from large skill ecosystems at inference time.
3/31/2026
Synthetic Mixed Training: Scaling Parametric Knowledge Acquisition Beyond RAG
Synthetic Mixed Training method combining synthetic QAs and documents to scale LLM knowledge acquisition beyond RAG limitations.
3/31/2026
Can LLMs Beat Classical Hyperparameter Optimization Algorithms? A Study on autoresearch
Study comparing LLM agents against classical hyperparameter optimization algorithms using autoresearch framework for hyperparameter tuning.
3/31/2026
PEANUT: Perturbations by Eigenvector Alignment for Attacking Graph Neural Networks Under Topology-Driven Message Passing
Attack method (PEANUT) exploiting GNN vulnerabilities through graph topology perturbations in message passing.
3/31/2026
Shapley meets Rawls: an integrated framework for measuring and explaining unfairness
Framework using Shapley values to measure and explain unfairness in ML models under group fairness criteria.
3/31/2026
Image-Adaptive GAN based Reconstruction
Image reconstruction using adaptive GANs to improve sample quality for complex image classes like human faces.
3/31/2026
Evasion Adversarial Attacks Remain Impractical Against ML-based Network Intrusion Detection Systems, Especially Dynamic Ones
Study on adversarial evasion attacks against ML-based network intrusion detection systems, showing attacks remain impractical against dynamic systems.
3/31/2026
LSM-GNN: Large-scale Storage-based Multi-GPU GNN Training by Optimizing Data Transfer Scheme
Multi-GPU GNN training system optimizing data transfer for large-scale graphs exceeding GPU memory capacity via storage-based approaches.
3/31/2026
NARVis: Neural Accelerated Rendering for Real-Time Scientific Point Cloud Visualization
Neural renderer for real-time visualization of large-scale scientific point cloud datasets using neural deferred rendering.
3/31/2026
Predictive variational inference: Learn the predictively optimal posterior distribution
Framework for variational inference that optimizes posterior distributions under model misspecification for better predictive accuracy.
3/31/2026
Trans-Glasso: A Transfer Learning Approach to Precision Matrix Estimation
Transfer learning method for precision matrix estimation leveraging related source data with limited target samples.
3/31/2026
Mean--Variance Portfolio Selection by Continuous-Time Reinforcement Learning: Algorithms, Regret Analysis, and Empirical Study
RL approach for continuous-time portfolio selection using unknown diffusion processes, with data-driven strategy learning.
3/31/2026
Efficient Human-in-the-Loop Active Learning: A Novel Framework for Data Labeling in AI Systems
Active learning framework optimizing expert labeling efficiency for training AI systems on unlabeled data.
3/31/2026
Synergizing Large Language Models and Task-specific Models for Time Series Anomaly Detection
Hybrid approach combining LLMs with task-specific models for time series anomaly detection, leveraging expert knowledge and pattern extraction.
3/31/2026
Hybrid Action Based Reinforcement Learning for Multi-Objective Compatible Autonomous Driving
RL approach for multi-objective autonomous driving that addresses policy updating and execution challenges in diverse driving scenarios.
3/31/2026
MaaS Updates
Platform offering unified API access to multiple open-source frontier AI models with optimized infrastructure for building agents and applications.
3/31/2026
Curious. anyone here allow agents to make purchase decisions of >$100?
Discussion asking whether anyone trusts AI agents to autonomously make purchase decisions over $100.
3/31/2026
Show HN: Gives your AI agents a shared, searchable, persistent memory – locally
SQLite-backed MCP memory server enabling multi-agent systems to share persistent, searchable memory with hybrid search locally.
3/31/2026
Adobe Illustrator can now use AI to rotate 2D vectors in 3D space
Adobe Illustrator feature enabling 2D vector illustrations to be rotated in 3D space while maintaining art style.
3/31/2026
Universal Claude.md – cut Claude output tokens by 63%
Drop-in configuration file that reduces Claude API output token usage by 63% without code changes.
3/31/2026
Whispr Flow – Vision Flow
Proof-of-concept multimodal interface using hand gestures and voice instead of text input for computer interaction.
3/31/2026
GTabs – AI tab organizer for Chrome that works with any LLM
Chrome extension using LLMs to automatically organize browser tabs into color-coded groups with review and undo capabilities.
3/31/2026
The AI Shift: Will software engineers survive agentic AI?
Paywall article speculating on impact of agentic AI on software engineers. No accessible content.
3/31/2026
Ask HN: After three years of open source software, I can't stand it
MarkFlowy note app with integrated AI (Copilot, DeepSeek, ChatGPT). Built with Tauri.
3/31/2026
Managing API keys and launching coding agents across providers
Aivo CLI manages API keys and launches coding agents across LLM providers. Supports Claude Code, local models, DeepSeek.
3/31/2026
Emerging Litigation Risks in Financing AI Data Centers Boom
Analysis of litigation risks in AI data center financing amid revenue-to-capex gap. Financial perspective.
3/31/2026
Geomys – a Gopher client for Macintosh 68K (2026)
Gopher browser for classic Macintosh 68K built with Claude Code. Retro computing project.
3/31/2026
Scaling Last-Mile route optimization to 1M stops on a laptop
Route optimization algorithm scaling to 1M stops. Limited details provided.
3/30/2026
Show HN: Claude/OpenAI/Gemini agents compete as investors with $100K each
Three LLM agents (Claude, GPT, Gemini) compete as virtual stock traders with $100K each using real market data. Demonstrates Upstash Box agent server primitive with isolated containers and autonomous tool usage.
3/30/2026
Ask HN: Are you too getting addicted to the dev workflow of coding with agents?
Discussion about using AI agents in development workflow. Developer shares experience of iterative agent-assisted coding loop.
3/30/2026
Show HN: Botference – A TUI to Plan with Claude Code and Codex Simultaneously
Terminal UI tool enabling collaborative project planning with Claude and Codex simultaneously in 'council' and 'caucus' modes. Outputs implementation plans for workflows.
3/30/2026
Show HN: AI for Your Team
Workspace platform enabling non-technical teams to collaborate with AI agents. Features memory, skills, MCP, scheduled tasks, and enterprise data integration with customization options.
3/30/2026
Freedom Studio – local AI runner with Tor, E2E encryption, and zero telemetry
Open-source local AI model runner with Tor integration, end-to-end encryption, and zero telemetry as privacy-focused alternative to LM Studio.
3/30/2026
Exo: Claude Code for your Inbox
Open source AI-native email client using Claude. Built with Electron, React, TypeScript. Analyzes and prioritizes emails with AI.
3/30/2026
Show HN: I built a Bitcoin signing device where the private key is physical
Hardware wallet design using physical titanium plate encoding private keys instead of persistent digital storage.
3/30/2026
Show HN: An Interactive Text to SQL Agent Benchmark
Benchmark for evaluating LLM models on text-to-SQL agent tasks, covering models from Opus to Qwen 0.8B with in-browser execution and visualizations.
3/30/2026
Pi: Another AI agent toolkit, but this one is interesting
Open-source toolkit for building AI agents and managing LLM deployments, includes coding agent package and contribution guidelines.
3/30/2026
The Blackwall Between Your AI Agent and Your Filesystem
Analysis of security permissions granted to AI coding agents and proposal for sandboxing mechanisms to restrict filesystem access without requiring Docker.
3/30/2026
PostMac – AI-native development, trunk-based Git, 12-factor philosophy
Personal development tools built with AI-native philosophy, trunk-based Git, and single-binary/HTML architecture.
3/30/2026
Bob Dylan's AI "Lectures from the Grave" Review: An Accidental Warning
Opinion piece reviewing Bob Dylan's AI-generated content on Patreon called 'Lectures from the Grave.'
3/30/2026
AI models sabotaging shutdown scripts. It took 22 years to regulate Meta
Opinion piece drawing parallels between Meta's 22-year regulatory lag and future AI model regulation.
3/30/2026
Ask HN: Do multiple short stints in startups hurt even if you've learned a lot?
Career advice question about hiring implications of multiple short startup tenures.