TSR · The desk

Compute

Articles in this category.

• TSR Desk · compute · 24 October 2026, 01:00 UTC

  What Pairwise Evaluation Misses: Reward Scores across Prompts

• TSR Desk · compute · 21 October 2026, 01:00 UTC

  Test-Time Compute Games

• TSR Desk · compute · 21 October 2026, 01:00 UTC

  BatchParScale: Sparse Batch-Mixing for Sequence Models

• TSR Desk · compute · 20 October 2026, 07:00 UTC

  Causal Scene Narration: Inference-Time Text Restructuring for Vision-Language-Action Driving

• TSR Desk · compute · 17 October 2026, 01:00 UTC

  When Does Muon Actually Win? A Paired Data-, Time-, and Tuning-Budget Benchmark

• TSR Desk · compute · 16 October 2026, 01:00 UTC

  ProbabilityBench: Model Evaluation and Calibration with Exact Ground-Truth Uncertainty

• TSR Desk · compute · 16 October 2026, 01:00 UTC

  When Log Parsing Removes What Matters: Detecting Failures in Distributed GPU Workloads

• TSR Desk · compute · 15 October 2026, 01:00 UTC

  FlowWise: Stateful Fast-Slow Model Serving for Streaming Traffic Intelligence

• TSR Desk · compute · 15 October 2026, 01:00 UTC

  Neural Bayesian Networks for Modular and Scalable Decision-Making Under Uncertainty

• TSR Desk · compute · 14 October 2026, 07:00 UTC

  Toward System One Clinical Intelligence: Parameter-Efficient Fine-Tuning and Counterfactual

• TSR Desk · compute · 13 October 2026, 01:00 UTC

  ThinkNet: Evolutionary Deep Ensembles for Efficient Motor-Imagery EEG Classification

• TSR Desk · compute · 13 October 2026, 01:00 UTC

  First-Order In-Context Shapley: Scalable Data Auditing for LLM Alignment and Evaluation

• TSR Desk · compute · 12 October 2026, 07:00 UTC

  Automatic-Differentiation-Enabled Finite Element Model Updating: Learning Constitutive Neural

• TSR Desk · compute · 9 October 2026, 01:00 UTC

  RLDISCOVER: LLM-driven co-evolution of reinforcement learning algorithms

• TSR Desk · compute · 8 October 2026, 01:00 UTC

  Teaching Agents to Code Reliably

• TSR Desk · compute · 8 October 2026, 01:00 UTC

  TasteVal: Measuring the Experimental Research Taste of AI Systems Against Human Experts

• TSR Desk · compute · 7 October 2026, 01:00 UTC

  Graph-Based Floor Separation Using Node Embeddings and Clustering of WiFi Trajectories

• TSR Desk · compute · 6 October 2026, 01:00 UTC

  How Far Can You Get Without a GPU? A Systematic Benchmark of Lightweight Hallucination

• TSR Desk · compute · 6 October 2026, 01:00 UTC

  Beyond Trained Models: Compiling GNNs for a Sound Explainer Benchmark

• TSR Desk · compute · 5 October 2026, 07:00 UTC

  D2K-Bench: Can LLM Agents Turn Expert Designs into Efficient GPU Kernels?

• TSR Desk · compute · 5 October 2026, 07:00 UTC

  NegT2IBench: When Negation Changes the Picture. A Polarity Benchmark for Text-to-Image Models

• TSR Desk · compute · 1 October 2026, 01:00 UTC

  Scaling Video Generation for Reasoning: At What Cost?

• TSR Desk · compute · 1 October 2026, 01:00 UTC

  Engineering Efficient Self-Play Chess: Search, Replay, and Throughput Under Limited Compute

• TSR Desk · compute · 1 October 2026, 01:00 UTC

  Agentic AI for Clustering, Relationship Discovery, and Semantic Trading in Prediction Markets

• TSR Desk · compute · 30 September 2026, 01:00 UTC · sooner

  AutoDataBench: Can Agents Write the Data That Feeds the Self-Improvement Loop?

• TSR Desk · compute · 30 September 2026, 01:00 UTC

  OneCanvas: 3D Scene Understanding via Panoramic Reprojection

• TSR Desk · compute · 29 September 2026, 01:00 UTC

  Efficient Safety Benchmarking via Item Response Theory

• TSR Desk · compute · 29 September 2026, 01:00 UTC

  Accounting for Bias Enables Sustainable LLM Evaluation

• TSR Desk · compute · 29 September 2026, 01:00 UTC

  ENAS: An Efficient Hardware-Aware Neural Architecture Search Framework for TinyML on

• TSR Desk · compute · 28 September 2026, 07:00 UTC

  Cartograph: Federated Tool Discovery with Operator-Attested Retrieval for AI Agents

• TSR Desk · compute · 28 September 2026, 07:00 UTC

  Energy Efficiency of Locally Deployed LLMs: A Preliminary Quantitative GPU Power Benchmark on

• TSR Desk · compute · 26 September 2026, 01:00 UTC

  Math Reasoning in LLMs is Organized by Approach, Not Topic

• TSR Desk · compute · 26 September 2026, 01:00 UTC

  How Reproducible Are Evaluation Conclusions? A Self-Audit of LLM-Inferred Prompt Structure

• TSR Desk · compute · 26 September 2026, 01:00 UTC

  RECLAIM: Can Agents Reproduce the Claims of Machine Learning Papers?

• TSR Desk · compute · 25 September 2026, 01:00 UTC

  SWE-Serve: Benchmarking Agentic Engineering For Production Inference Serving

• TSR Desk · compute · 24 September 2026, 01:00 UTC

  Efficient Iterative Retrieval with Heterogeneous Batching

• TSR Desk · compute · 23 September 2026, 01:00 UTC

  How Much of a Real Workload Can LLM-Generated GPU Kernels Actually Reach?

• TSR Desk · compute · 23 September 2026, 01:00 UTC

  Bio-MF: Low-Latency and High-Fidelity EEG-to-fNIRS Cross-Modal Generation for Hybrid

• TSR Desk · compute · 23 September 2026, 01:00 UTC

  GeoSelect: Spatial-Program Execution for Training-Free Referring Remote Sensing Image

• TSR Desk · compute · 22 September 2026, 01:00 UTC

  SynthDemo-RL: Breaking the Zero-Reward Barrier in VLA Adaptation with LLM-Guided Synthetic

• TSR Desk · compute · 21 September 2026, 07:00 UTC

  Verify, Don't Trust: Agentic Model Development for Video Discovery Retrieval at Scale

• TSR Desk · compute · 21 September 2026, 07:00 UTC

  Scaling Discovery through Test-Time Communication

• TSR Desk · compute · 19 September 2026, 01:00 UTC

  Predict Before You Deploy: Offline Prediction of Quantization-Induced Task Degradation for

• TSR Desk · compute · 19 September 2026, 01:00 UTC

  Self Improvement via Fast Tree-search

• TSR Desk · compute · 19 September 2026, 01:00 UTC

  Beyond Depth Truncation: Controlled Evaluation of Depth Utilization in Recursive Language Models

• TSR Desk · compute · 18 September 2026, 01:00 UTC

  One Color Preprocessing Improves DSATUR

• TSR Desk · compute · 18 September 2026, 01:00 UTC

  Where Should Agents Live? Energy-Memory Characterization of Agentic AI for the Edge-Cloud

• TSR Desk · compute · 18 September 2026, 01:00 UTC

  When AI Generates Covariates: Causal Typing and Estimand Drift in Sequential Experiments

• TSR Desk · compute · 18 September 2026, 01:00 UTC

  Ask the Tool, Don't Guess: Agent Tool Calls Hold Their Progress, and the Serving System Should

• TSR Desk · compute · 16 September 2026, 01:00 UTC

  LLMs or Naive Bayes? Old Gems or New Ways

• TSR Desk · compute · 16 September 2026, 01:00 UTC

  Echo-CoPilot: A Multiple-Perspective Agentic Framework for Reliable Echocardiography

• TSR Desk · compute · 16 September 2026, 01:00 UTC

  MemRiskBench: Trace-Aware Risk-Preserving Evaluation for Long-Horizon LLM Agents

• TSR Desk · compute · 14 September 2026, 07:00 UTC

  Countdown-Code: A Testbed for Studying The Emergence and Generalization of Reward Hacking in

• TSR Desk · compute · 14 September 2026, 07:00 UTC

  Evidence-Unit Fairness and the Limits of Query-Adaptive Sparse-Dense Fusion in Financial

• TSR Desk · compute · 13 September 2026, 01:00 UTC

  Towards a Deterministic Math Solver for Clinical Language Models

• TSR Desk · compute · 13 September 2026, 01:00 UTC

  An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics

• TSR Desk · compute · 12 September 2026, 01:00 UTC

  Programmable Cellular Automata

• TSR Desk · compute · 11 September 2026, 01:00 UTC

  Compute-Bounded Security Assurance - Coverage, Verification, and Response under Resource

• TSR Desk · compute · 11 September 2026, 01:00 UTC

  Proof-Carrying Cognition: Closing the Verification Gap with Reality-Settled Reward

• TSR Desk · compute · 10 September 2026, 01:00 UTC

  Jetson-PI: Towards Onboard Real-Time Robot Control via Foresight-Aligned Asynchronous Inference

• TSR Desk · compute · 9 September 2026, 07:00 UTC

  Bait-and-Recover: Poisoning Internal Refusal Signals to Defend LLMs against White-Box Editing

• TSR Desk · compute · 9 September 2026, 07:00 UTC

  AtomCite: Verification and Correction of Supplied Page-Level Citations in Multi-Page Documents

• TSR Desk · compute · 8 September 2026, 01:00 UTC

  Substrate-Aware AI Agents: Execution Context as a First-Class Input

• TSR Desk · compute · 7 September 2026, 07:00 UTC

  Iris: Climbing to the Search Frontier

• TSR Desk · compute · 4 September 2026, 19:01 UTC

  It's the Problem, Not the Path: Budget and Difficulty Confounds in LLM Reasoning Trajectories

• TSR Desk · compute · 4 September 2026, 19:01 UTC

  Xiaomi-TabLDM: A Tabular Foundation Model Technical Report

• TSR Desk · compute · 4 September 2026, 19:01 UTC

  A Comparative Study in Surgical AI: Potential and Limitations of Data, Compute, and Scaling

• TSR Desk · compute · 4 September 2026, 19:01 UTC

  Verify Before You Distill: Prompt-Level Teacher Gating for On-Policy Distillation

• TSR Desk · compute · 4 September 2026, 19:01 UTC

  TAP-Path: Task-Adaptive Structural and Token Pruning for Efficient and Trustworthy Pathology

• TSR Desk · compute · 4 September 2026, 19:01 UTC

  Free Pause Tokens

• TSR Desk · compute · 4 September 2026, 19:01 UTC

  Sionna RT: Technical Report