2/17/2026
Enhancing Delta Compression in LLMs via SVD-based Quantization Error Minimization
SVD-based quantization for compressing fine-tuned LLM delta parameters with minimized quantization error for storage efficiency.
2/17/2026
HYPER: A Foundation Model for Inductive Link Prediction with Knowledge Hypergraphs
HYPER foundation model for inductive link prediction with knowledge hypergraphs handling novel entities and relation types.
2/17/2026
Calibrated Predictive Lower Bounds on Time-to-Unsafe-Sampling in LLMs
Novel safety measure for LLMs measuring generations needed to trigger unsafe responses, with calibrated predictive bounds for evaluation.
2/17/2026
NeuronSeek: On Stability and Expressivity of Task-driven Neurons
NeuronSeek uses symbolic regression to discover optimal neuron formulations and construct task-driven neural networks.
2/17/2026
Vision Transformers for Multi-Variable Climate Downscaling: Emulating Regional Climate Models with a Shared Encoder and Multi-Decoder Architecture
Vision Transformers for climate downscaling using shared encoder and multi-decoder architecture to emulate regional climate models.
2/17/2026
Chain of Thought in Order: Discovering Learning-Friendly Orders for Arithmetic
Studies optimal ordering of intermediate steps in chain-of-thought reasoning for arithmetic tasks, improving learning efficiency.
2/17/2026
wd1: Weighted Policy Optimization for Reasoning in Diffusion Language Models
Weighted policy optimization method for reinforcement learning in diffusion-based LLMs, addressing intractable likelihood approximation.
2/17/2026
The Serial Scaling Hypothesis
Theoretical framework distinguishing inherently sequential problems that cannot be efficiently parallelized, relevant to LLM reasoning.
2/17/2026
DeepC4: Deep Conditional Census-Constrained Clustering for Large-scale Multitask Spatial Disaggregation of Urban Morphology
Spatial disaggregation clustering method for large-scale urban morphology mapping using Earth observation data.
2/17/2026
FGBench: A Dataset and Benchmark for Molecular Property Reasoning at Functional Group-Level in Large Language Models
Benchmark dataset for evaluating LLM reasoning on molecular properties at functional group level for chemistry applications.
2/17/2026
Beyond Prompt-Induced Lies: Investigating LLM Deception on Benign Prompts
Investigates spontaneous deception in LLMs on benign prompts, revealing trustworthiness risks in reasoning and planning tasks.
2/17/2026
Lightning Prediction under Uncertainty: DeepLight with Hazy Loss
Deep learning architecture for lightning occurrence prediction using uncertainty-aware loss function.
2/17/2026
MAVIS: Multi-Objective Alignment via Inference-Time Value-Guided Selection
Inference-time method for balancing multiple conflicting objectives in LLM outputs without expensive per-objective fine-tuning.
2/17/2026
Predicting the Order of Upcoming Tokens Improves Language Modeling
Token order prediction auxiliary objective improves language model performance, offering alternative to multi-token prediction for next-token training.
2/17/2026
ART: Adaptive Resampling-based Training for Imbalanced Classification
Adaptive resampling method for imbalanced classification that adjusts training data distribution based on class-wise learning difficulty.
2/17/2026
Online reinforcement learning via sparse Gaussian mixture model Q-functions
Online policy-iteration RL framework using sparse Gaussian mixture model Q-functions with interpretable exploration mechanisms.
2/17/2026
DiffusionNFT: Online Diffusion Reinforcement with Forward Process
Online reinforcement learning method for diffusion models addressing intractable likelihoods, enabling RLHF-style training without solver restrictions.
2/17/2026
Functional Scaling Laws in Kernel Regression: Loss Dynamics and Learning Rate Schedules
Analyzes scaling laws for loss dynamics and learning rate schedules in SGD on kernel regression, with implications for LLM training.
2/17/2026
Learning the Inverse Temperature of Ising Models under Hard Constraints using One Sample
Theoretical study on estimating inverse temperature parameters in truncated Ising models using single samples.
2/17/2026
The Rogue Scalpel: Activation Steering Compromises LLM Safety
Demonstrates that activation steering for LLM control can compromise safety mechanisms, causing models to comply with harmful requests.
2/17/2026
Effective Quantization of Muon Optimizer States
8-bit quantization technique for Muon optimizer states in LLM pre-training, reducing memory overhead while maintaining training efficiency.
2/17/2026
Better Hessians Matter: Studying the Impact of Curvature Approximations in Influence Functions
Studies how curvature approximations (GGN, K-FAC) in influence functions affect data attribution accuracy for deep learning models.
2/17/2026
LLM DNA: Tracing Model Evolution via Functional Representations
Method to trace evolutionary relationships between LLMs through functional representations, enabling better model management and understanding of fine-tuning/distillation lineages.
2/17/2026
OpenTSLM: Time-Series Language Models for Reasoning over Multivariate Medical Text- and Time-Series Data
OpenTSLM: time-series language models integrating multivariate medical time-series as native modality. Enables LLMs to handle temporal clinical data.
2/17/2026
Multi-scale Autoregressive Models are Laplacian, Discrete, and Latent Diffusion Models in Disguise
arXiv paper: Visual Autoregressive models reinterpreted as Laplacian latent pyramid with learned coarse-to-fine refinement. Formal analysis of design trade-offs.
2/17/2026
Where to Add PDE Diffusion in Transformers
Research on optimal placement of PDE diffusion layers in transformer architectures using heat equation-based smoothing for local geometric priors.
2/17/2026
SciTech SNAP Graphics has gone open source
Legacy SciTech SNAP Graphics device driver codebase released as open source.
2/17/2026
Show HN: M-Courtyard – Fine-tune LLMs on your Mac with zero code
Zero-code desktop app for fine-tuning LLMs locally on Apple Silicon with full pipeline from documents to Ollama export.
2/17/2026
Show HN: Broadcast Box – Self-hosted low latency streaming
Self-hosted low-latency streaming application using WebRTC for sub-second broadcast.
2/17/2026
Show HN: ExoGen – Open-Source Local Stable Diffusion Client
Privacy-focused desktop app for local Stable Diffusion image generation with HuggingFace model support.
2/17/2026
DelegateOS: Cryptographic delegation tokens for multi-agent sys(DeepMind paper)
DeepMind's cryptographic delegation protocol for multi-agent systems with accountability chains and MCP integration.
2/17/2026
Almost Every infrastructure decision I endorse or regret after 4 years
Retrospective on infrastructure decisions at startup including cloud provider choice, database selection, and tradeoffs.
2/17/2026
LLM-generated skills work, if you generate them afterwards
Research showing LLM-generated skills provide no benefit; models cannot reliably author procedural knowledge they consume.
2/17/2026
Manus AI launched 24/7 Agent via Telegram and got suspended
Manus AI launches persistent AI agents with Telegram integration; account suspended shortly after launch. Platform expansion planned via WhatsApp.
2/17/2026
Show HN: Purely Vibe Coded Asmongold Simulator
Vibe-coded Asmongold simulator game with author expressing skepticism about AI-based code generation.
2/17/2026
Agent-evals: Overlap, boundary, and metacognitive scoring for coding agents
Evaluation framework for coding agents detecting overlap, boundary violations, and coverage gaps via static and live analysis.
2/17/2026
Show HN: The first financial intelligence MCP server live trading signals Claude
MCP server providing Claude real-time access to trading signals from Reddit, SEC filings, FDA approvals, and Congressional trades.
2/17/2026
Show HN: Forage – MCP server that lets AI agents find and install their own MCPs
MCP server enabling AI agents to discover, install, and learn new tools automatically without restarts or manual configuration.
2/17/2026
Show HN: Constrained DSL for Reliable LLM Decisions
Constrained DSL for generating reliable LLM decision logic with schema-driven prompts and deterministic execution for quantitative tasks.
2/17/2026
Show HN: An Open-source React UI library for ASCII animations
React library for ASCII animations that converts video to character grids with performance optimization.
2/17/2026
Shard – A Distributed P2P AI Network for Shared Inference
Minimal title-only entry about distributed P2P network for AI inference without substantive content.
2/17/2026
Meta patented an AI that lets you keep posting from beyond the grave
Meta patent allowing LLM simulation of deceased users' social media activity for continued posting.
2/17/2026
Game developers and pixel artists are losing their jobs
Marketing content for generative sprite creation tool for game developers, claiming to replace artist jobs.
2/17/2026
India's 'AI Impact Summit' Promises Little More Than Spectacle
Opinion on India's AI policy summit with focus on governance and homelessness concerns.
2/17/2026
Beating GPT-2 for less than $100 – Andrej Karpathy
Andrej Karpathy explores training language models competitive with GPT-2 for under $100, analyzing cost-effectiveness improvements since 2019.
2/17/2026
Show HN: Bulwark – Open-source governance layer for AI agents (Rust, MCP-native)
Open-source governance proxy in Rust for controlling AI agent access to tools, providing audit trails and content moderation via MCP.
2/17/2026
Why I Built Reader: Open-source web scraping for LLMs
Open-source web scraper optimized for LLM consumption, cleaning HTML noise for agent accessibility to web content.
2/17/2026
New GitHub repository settings to configure pull request access
GitHub repository settings update allowing maintainers to disable or restrict pull requests. Not AI/ML related.
2/17/2026
GrowthClaw, Distribution Infrastructure for OpenClaw
GrowthClaw is an open-source marketing operating system for agent-driven workflows, converting goals into task pipelines with human approval and evaluation.
2/17/2026
AI is destroying Open Source, and it's not even good yet
Opinion piece critiquing AI's impact on open source, discussing hallucinations, agent harassment, and OpenAI recruitment.