Stories by matt_d

Learning to Repair Lean Proofs from Compiler Feedback

RLix: A scheduling layer for concurrent LLM RL

Primus Projection: Estimate Memory and Performance Before You Train

PRowhammer: Propagating Bit-flips from CPU to GPU [pdf]

The Quantization Robustness of Diffusion Language Models in Coding Benchmarks

Different Perspectives of Memory System Simulation

Adding Compilation Metadata to Binaries to Make Disassembly Decidable

ICLR 2026 Outstanding Papers

Decoupled DiLoCo for Resilient Distributed Pre-Training

spmd_types: A type system for distributed (SPMD) tensor computations in PyTorch

How Do LLM Agents Think Through SQL Join Orders?

Gluon&Linear Layouts Deep-Dive:Tile-Based GPU Programming with Low-Level Control [video]

SonicMoE: A HW-Efficient and SW-Extensible Blueprint for Fine-Grained MoEs

SAW-INT4: System-Aware 4-Bit KV-Cache Quantization for Real-World LLM Serving

DIRT: Database-Integrated Random Testing

Scaling Test-Time Compute for Agentic Coding

An Algorithmic Reconstruction of Normalisation by Evaluation

Faster LLM Inference via Sequential Monte Carlo

Pure Borrow: Linear Haskell Meets Rust-Style Borrowing

SSA without Dominance for Higher-Order Programs

7 8 9 10 11 12 13 14 15 16