HN talvardi7 27d ago

Switching AI Tools Mid-Sprint Cost Us a Day (and What We Learned) – Week 6 R

Case study on switching AI tools mid-project causing productivity loss. Brief anecdote without technical depth.

HN kris_osiadacz 27d ago

Why local AI – and why it matters

Argument for local, on-premise AI systems vs cloud AI for privacy and reliability. Conceptual analysis of deployment approaches.

HN nyellin 27d ago

How to Drive an LLM

Guide on LLM prompting and interaction techniques. Incomplete content.

HN maarcel93 27d ago

What 'Getting Your Hands Dirty' Means at LLM-Era

Article about practical LLM engineering and hands-on development practices. Incomplete content.

HN LeoPanthera 27d ago

Gemini provides phone number of scammer posing as Delta Airlines

User scammed using information from Gemini chatbot. Anecdotal report on LLM providing unreliable information.

HN tyxak 27d ago

RemotePower – self-hosted fleet monitoring with built-in vulnerability scanning

RemotePower is self-hosted Linux fleet monitoring with CVE scanning, RAG-based documentation search, and optional AI integration.

HN giamma 27d ago

High-performance code intelligence MCP server

High-performance MCP server provides fast code intelligence for AI agents, indexing large repositories in seconds with tree-sitter AST parsing across 158 languages.

HN rkdgh19 27d ago

Show HN: Redteam:If you are using more than 2 coding agents

Redteam is an adversarial agent-pair harness for AI-assisted coding that uses two independent models to review and challenge code outputs before merge.

HN matt_d 27d ago

Dana Scott: Lambda Calculus, Forcing and the Foundations of Math: #14 aboutlogic [video]

Video lecture on lambda calculus, forcing, and mathematical foundations by Dana Scott.

HN theanonymousone 27d ago

I tested Gemma4 12B on my 8GB GPU, now I don't want to go back to smaller models

User tests Google's Gemma4 12B local LLM on 8GB GPU, finding recent model improvements focus on consumer hardware efficiency over scale.

Ax Animesh Maheshwari, Divyansh Sahu, Nishit Verma 27d ago

Do Vision-Language Models Understand 3D Scenes or Just Catalogue Objects?

Benchmark evaluating whether vision-language models understand 3D spatial layout beyond object recognition via occlusion and geometry tasks.

Ax Slim Barkallah, Luke Bailey, Kaiyue Wen, Mohammed Abouzaid, Tengyu Ma 27d ago

Pseudo-Formalization for Automatic Proof Verification

Pseudo-Formalization method automatically converts AI-generated mathematical proofs into verifiable formal language representations.

Ax Ismail Geles, Leonard Bauersfeld, Markus Wulfmeier, Davide Scaramuzza 27d ago

Superhuman Safe and Agile Racing through Multi-Agent Reinforcement Learning

Multi-agent reinforcement learning approach for safe, superhuman autonomous racing coordination in dynamic shared spaces.

Ax Jiacheng Lu, Haoyi Zhu, Sipei Yi, Enze Xie, Yu Li, Cheng Zhuo 27d ago

Light Interaction: Training-Free Inference Acceleration for Interactive Video World Models

Light Interaction provides training-free inference acceleration for interactive video world models via optimized attention and caching.

Ax Xinming Wei, Chao Jin, Tuo Dai, Yinmin Zhong, Shan Yu, Chengxu Yang, Bingyang Wu, Zili Zhang, Jing Mai, Qianchao Zhu, Zhouyang Li, Yuliang Liu, Guojie Luo 27d ago

UltraEP: Unleash MoE Training and Inference on Rack-Scale Nodes with Near-Optimal Load Balancing

UltraEP system optimizes expert parallelism training and inference for mixture-of-experts models with dynamic load balancing.

Ax Nishant Subramani, Palash Goyal, Yiwen Song, Mani Malek, Yuan Xue, Tomas Pfister, Hamid Palangi 27d ago

The ACUTE Protocol: Operationalizing Language Model Activations for Better Calibration, Utility, and Trust

ACUTE Protocol uses language model internal activations to improve confidence calibration, reducing overconfidence and enhancing trustworthiness.

Ax Apratim Bhattacharyya, Shweta Mahajan, Sanjay Haresh, Rajeev Yasarla, Reza Pourreza, Litian Liu, Risheek Garrepalli, Roland Memisevic 27d ago

Streaming Interventions: Can Video Large Language Models Correct Mistakes as They Occur?

Evaluates video large language models' ability to provide real-time mistake correction guidance during instructional task execution.

Ax Spyros Rigas, Ioannis Contopoulos, Georgios Alexandridis, Antonios Nathanail 27d ago

An adaptive framework for the axisymmetric pulsar magnetosphere using physics-informed Kolmogorov-Arnold networks

Physics-informed Kolmogorov-Arnold networks with domain-specific adaptation for modeling axisymmetric pulsar magnetosphere dynamics.

Ax Jian Xu, Chao Li, Delu Zeng, John Paisley, Qibin Zhao 27d ago

Higher-Order Token Interactions via Quantum Attention

Quantum Higher-Order Attention mechanism for computing higher-order token interactions beyond pairwise relationships in transformers.

Ax Batu Candan, Simone Servadio 27d ago

Deep Learning-Based Lunar Crater Terrain Relative Navigation

Deep learning crater detection combined with Extended Kalman Filter for terrain relative navigation during lunar landings.

Ax Qing Huang, Pooja Pol, Jianing Zhang 27d ago

LLM-Based Synthetic Ground Truth Generation for Audio-Based Emotion Classification via In-Context Learning

Uses LLMs via in-context learning to generate synthetic ground truth labels for training audio-based emotion classification models.

Ax Oph\'elia Miralles, M\'at\'e Mile, Christoffer Artturi, Thomas Nipen, Ivar Seierstad 27d ago

Pointwise is Pointless? A Multimodal Ablation Study for Precipitation Nowcasting with Graph Neural Networks

Multimodal graph neural network system for precipitation nowcasting combining radar, weather prediction, and sparse point observations.

Ax Shogo Yamauchi, Hideaki Tamori, Makoto Sakai, Yosuke Yamano, Tohru Nitta 27d ago

QC-GAN: A Parameter-Efficient Quaternion Conformer GAN for High-Fidelity Speech Enhancement

QC-GAN proposes parameter-efficient speech enhancement model combining Quaternion Conformer with metric-learning discriminator training.

Ax Christopher En, Yuri Faenza, Andrea Lodi, Gonzalo Mu\~noz 27d ago

Fair Online Resource Allocation

Theoretical study of fair online resource allocation algorithms for sequential agent assignment with fairness constraints and capacity limits.

Ax Nahum Korda, Gadi Evron 27d ago

OpenAnt: LLM-Powered Vulnerability Discovery Through Code Decomposition, Adversarial Verification, and Dynamic Testing

OpenAnt framework uses LLMs for automated vulnerability discovery in code repositories via semantic reasoning, decomposition, and adversarial verification.

Ax Mengxiang Hao, Xin Jiang, Xinghao Huang, Wenliang Su, Zhiteng Wang, Junjie Rao, Xiaotian Yang, Wei Liao, Chengyu Han, Gen Liang, Yulun Song, Zhitao Xu, Xianpeng Lang 27d ago

Learning to Annotate Delayed and False AEB Events: A Practical System for Extreme Class Imbalance and Asymmetric Label Noise

Automated annotation framework using machine learning to handle extreme class imbalance in Autonomous Emergency Braking system event labeling.

Ax Hannah Le, Ramesh Ramasamy, Alex Urrutia, Mahsa Yazdani, Tim Proctor, Kenny Workman 27d ago

TxBench-PP: Analyzing AI Agent Performance on Small-Molecule Preclinical Pharmacology

TxBench-PP benchmark for evaluating AI agents on small-molecule preclinical pharmacology tasks in drug discovery with verifiable, realistic decision-making scenarios.

HN qainsights 27d ago

Show HN: Iamspeed.dev – Fast.com Style but for LLMs

Iamspeed.dev is a speed testing tool for LLMs, similar to Fast.com but measuring language model performance.

HN jonbaer 27d ago

Agent Finder

Agent Finder: search interface implementing Agentic Resource Discovery specification; catalog of Azure skill definitions for agent integration.

HN Bender 27d ago

Everyone got replaced [video][23 mins]

Video content without meaningful description or technical details provided.

HN caudena 27d ago

Show HN: BeamWeaver – LangChain/DeepAgents-style agents and workflows for Elixir

BeamWeaver: Elixir framework for building agents and agentic workflows inspired by LangChain/LangGraph. Features tool calling, graph workflows, checkpoints, memory, and provider adapters.

HN A19dammer91 27d ago

Show HN: Mrs-Hybride-PQC – Hybrid Kyber1024 KEM 5-6x Faster Than HKDF-SHA256

Mrs-Hybride-PQC: hybrid post-quantum cryptography library combining Kyber1024 KEM with MRS Diophantine key derivation for 5-6x faster encryption than HKDF-SHA256.

HN Raj_Sidwadkar 27d ago

Finally MCP's Tool Poisoning Gap Solved: A Protocol-Level Defense

Protocol-level defense against MCP tool poisoning. Proposal and reference implementation for optional cryptographic signing of tool manifests to prevent modification after approval.

HN doppp 27d ago

Datasette-apps: Apps that live inside Datasette

Datasette-apps: plugin enabling custom HTML/JavaScript applications to run in iframe sandbox within Datasette, with read/write SQL query support.

HN lumpa 27d ago

Datasette Apps: Host custom HTML applications inside Datasette

Datasette Apps plugin launches, enabling custom HTML/JavaScript applications in sandboxed iframes that can query and write to Datasette databases.

HN javatuts 27d ago

How to Build ChatGPT from Scratch: Understanding LLMs Step by Step

Educational article explaining how LLMs work, covering terminology and practical applications in coding, document summarization, and autonomous agents.

HN tobypadilla 27d ago

Show HN: Flashback – an agent skill that references 127 years of design trends

Flashback: agent skill providing searchable historical design trends database from 1900-present as reference resource for design-related agent tasks.

HN sricola 27d ago

Show HN: Drydock – VM Sandboxes for macOS Autonomous Coding Agents

Drydock: VM sandbox system for running autonomous coding agents (Claude, OpenAI) locally on macOS with hardware isolation, token gating, and deny-by-default egress.

HN Psychoterapist 27d ago

Can LLMs change the hiring economics of legacy engineers?

Case study testing whether candidates with outdated tech stacks but active AI tool usage can pass standard hiring processes at software companies.

HN thedreammachine 27d ago

Ask HN: How much do you trust LLMs with your health questions?

Discussion thread asking software engineers about using LLMs for health advice compared to coding tasks, noting the difficulty of verification outside engineering domains.

HN osaeld 27d ago

Get user requests directly on your website and have an agent build

Agent-powered code modification system. Users request changes on live websites; coding agent implements changes, generates preview, and creates pull request for approval.

HN _zendar_ 27d ago

Show HN: Hermzner – Provisioning an Hardened Hermes Agent on Hetzner VPS

Hermzner: provisioning tool for deploying Hermes Agent on Hetzner VPS with security hardening, Tailscale integration, backups, and optional Mnemosyne memory provider.

HN rmason 27d ago

OpenMontage the first open-source, agentic video production system

OpenMontage: open-source agentic system for video production. Users describe desired output in natural language; agent handles research, scripting, asset generation, editing, and composition.

HN jankdc 27d ago

Show HN: bote - A fast, modern and low-memory approach to processing a big JSON

bote library processes large JSON files with low memory footprint using cursor-based streaming and path queries.

HN dpola 27d ago

Zork name origin got an update on Wikipedia

Article about Zork game name etymology and Wikipedia accuracy.

HN amaldavid 27d ago

Show HN: Pagecast – Publish Markdown/HTML Reports to Cloudflare Pages

Pagecast CLI tool for publishing Claude Code/Codex-generated HTML/Markdown reports to Cloudflare Pages.

LB elastic.co via Mordo 27d ago

Agent memory on Elasticsearch: hybrid retrieval and DLS

Elastic Agent Builder provides persistent memory layer for agents using hybrid retrieval, reranking, and decay mechanisms.

HN Divisibly3 27d ago

Glados for Claude

GLaDOS personality plugin with neural TTS for Claude Code using Forward Tacotron and HiFiGAN.

HN matt_d 27d ago

Integer Quantization: Deep Dive

Deep dive on integer quantization for transformers, covering techniques for fitting large models in 4-bits on single GPU.

HN posnetes 27d ago

Hacked POS terminals as live instruments and Open source browser interfaces

Minimal content about hacked POS terminals and browser interfaces.