makecindy
Featured
cindy
Cindy is an open-source AI agent that runs locally on your machine, integrates multiple AI harnesses and models, and provides memory, skills, and automation to perform real work in your projects and apps.
Hyra-results
This repository contains the concrete solutions and reproducible scripts produced by the Hyra (Hunyuan Research Agent) across a range of open problems in science, mathematics, engineering, and creative design, with most results surpassing the previous best published values.
Stars
100
7-day growth
+6
Forks
7
Open issues
0
License
NOASSERTION
Last updated
2026-07-27
AI repository intelligence
FR-AI / ANALYSISWhy it is worth attention
It demonstrates the capabilities of a research agent that outperforms prior state-of-the-art on multiple diverse benchmarks (AI4AI, AI4Science, and creative AI4Fun tasks), providing reproducible artifacts that allow the community to verify and build upon its results.
Who it is for
- AI researchers evaluating autonomous research agents
- Scientists and mathematicians looking for improved solutions to specific open problems
- Machine learning engineers interested in reproducible benchmarking
- Hobbyists exploring AI-generated creative designs (games, music, 3D models)
Use cases
- Benchmarking the performance of research agents against human-designed baselines
- Reproducing state-of-the-art results for tasks like autocorrelation constants, packing records, and symbolic forecasting
- Integrating Hyra's solutions as starting points for further optimization in continuously-updated leaderboards
- Studying AI-generated creative works such as Reversi bots, music arrangements, and 3D procedural models
Strengths
- Outperforms previous best published results on 11 out of 12 compared tasks across AI4AI and AI4Science
- Covers a wide variety of domains (AI research, physics, combinatorics, drug docking, quantum computing, creative design) demonstrating versatility
- Provides self-contained scripts and final solution artifacts for reproducibility and verification
- Includes notable creative demos (e.g., AlphaZero-style Reversi bot, music arrangement, 3D models) that showcase the agent's capabilities beyond pure optimization
Considerations
- Results are frozen as of 2026-07-10; later leaderboard entries may have improved upon Hyra's solutions
- The repository only contains the output artifacts, not the Hyra agent itself, so users cannot directly run the agent or modify its behavior
- Some tasks (e.g., AI4Fun) lack quantitative comparisons, making it harder to assess the agent's creative performance objectively
README quick start
Hyra Results
Companion artifacts for Hyra: Hnyuan Research Agent.
This repository collects concrete solutions that Hyra produced across a range of open problems in science, mathematics, engineering, and creative design, released alongside the Hyra launch post. Each folder holds the final solution artifact and, where relevant, the self-contained scripts that reproduce it.
📝 Launch post: hy.tencent.com/research/hyra
Results
Each row compares the best prior published result (Prev best, from the cited system or leaderboard) against Hyra. Arrows mark the better direction (↓ lower is better, ↑ higher is better); the Hyra-winning value is in bold.
| Track | Task | Metric | Prev best | Hyra |
|---|---|---|---|---|
| AI4AI | nanochat_autoresearch | val BPB ↓ | 0.9109 e | 0.9015 |
nanogpt_speedrun | wall-clock ↓ | 77.5 s e | 76.4 s | |
sol_execbench | score ↑ | 0.754 e | 0.771 | |
| AI4Science | autocorrelation_first | C₁ ↓ | 1.502870 a | 1.502850 |
autocorrelation_second | R ↑ | 0.962694 b | 0.962901 | |
erdos_min_overlap | C₅ ↓ | 0.380868 b | 0.380859 | |
sums_diffs | C(A) ↑ | 1.14489 b | 1.21079 | |
packing_records | records broken | n/a f | 100 | |
smallest_adder | params ↓ | 36 c | 15 | |
parp1_docking | objective ↓ | −9.77 d | −10.60 | |
qubit_routing | CNOTs added ↓ | 269,037 b | 258,369 | |
sunspot_symbolic | forecast R² ↑ | 0.47 g | 0.78 |
Prev-best sources.
- a TTT-Discover: Learning to Discover at Test Time (arXiv:2601.16175).
- b SimpleTES: Evaluation-driven Scaling for Scientific Discovery (arXiv:2604.19341).
- c AdderBoard trained-weights leaderboard (github.com/anadim/AdderBoard).
- d Olaparib, an approved PARP1 inhibitor (drug baseline).
- e Recursive: *First Steps Toward Automated AI
Description
Research artifacts from Hyra (/ˈhaɪ.rɑː/)
Related repositories
Similar projects matched by category, topics, and programming language.
deerwork-ai
Featured
deer-workflow
An open-source Dynamic Workflow runtime that combines deterministic TypeScript orchestration with replaceable Agent runtimes.
7-e1even
Featured
learn-agent
A collection of notes on building coding agents, derived from the production development of the desktop agent Reina, with each mechanism simplified into a zero-dependency, single-file Node.js demo.