Havuz

SWE-sweep, agent'ların büyük kod tabanlarındaki konumu ve türü önceden belirtilmeyen birden fazla bug'ı bulup düzeltebilmesini ölçen bir benchmark'tır.

Neden yükseliyor · 1 günlük, 33 yıldızlı yeni bir Facebook Research benchmark'ı olması ve LLM agent değerlendirmesine yönelik güncel bir soruyu ele alması dikkat çekebilir. Ancak yükselişin kesin nedeni bilinmiyor; son 28 günlük commit ve farklı yazar verisi ölçülmemiş.

[Public preview] Externally scored agentic ML research benchmark: 60 tasks, real competition ground truth. Open protocol, operated evaluatio

Track 42,331+ AI benchmark, eval, dataset, and data-quality records from 40 public sources, with linked evidence and daily updates.

Runtime evidence that helps agents trace, profile, and burn down hotspots in application and native code, GPU kernels, and inference stacks.

Community registry of LLM serving-path traps that produce confidently wrong measurements: templates, tool parsers, reasoning fields, quant k

A recursive self-improvement (RSI) workflow that evolves both target agents (e.g., Pi and DeepSeek Harness) and the RSI system itself.

CLI Noise-aware CPU pinning

A decision library for running local LLMs on multi-GPU machines without NVLink: parallelism, communication, quantization, speculative decodi

HEWN 2.0 2026: AI Output Router for Precision Summaries & Polished Code

Occam's razor for Claude Code and Codex: a 30-line rule set. Benchmarks vs base: 51-55% lower estimated task cost on Opus/Sonnet 5.5 max wit

Specification-driven multi-agent systems with OpenTelemetry observability, agent traces, benchmarking, MCP, and A2A.

Local workload scheduler for autonomous agents sharing an Apple Silicon Mac.

Semantic cache strategy bench: replay recorded request traces through candidate caching policies and measure which one actually pays off.

Cross-platform .NET performance engineering skill for coding agents, covering CPU, memory, GC, benchmarking, concurrency, startup, native pr

Iterative agent harness improvement: run a coding agent on a hard task, generate the reusable tooling it was missing, qualify it, and replay

Is your project as good as the best out there? sota-scan benchmarks your repo against real top projects and hands you a scorecard + to-do li

Benchmarks for pdf redaction tools

AMBER (sealed-in-amber historical replay evaluation): real-event replay, physical sealing, pre-registered scoring — method spec; cases priva

Unified Python/CLI time-series decomposition library with native acceleration, multivariate workflows, and machine-readable outputs.

Open browser arena for embodied / VLA stacking benchmarks

Browser-native, GPU-accelerated editor for very large hierarchical 2D IC-layout scenes, written in Rust (native and WebAssembly).

Run a command and report its process tree's CPU, memory, and I/O usage

Automated auditing pipeline for LLM and agent benchmarks — surfaces task ambiguity, environment conflicts, and evaluation bugs.

The open testing standard for voice AI agents. Deterministic + semantic + RAG augmented evaluation. Local first. Zero telemetry.