THE INTERNET IS THINKING
Earlier sources: Artificial Intelligence
Artificial intelligence covers systems that learn from data, generate content, perceive their surroundings, or act toward a goal. Follow model research, agent tooling, evaluation, and the infrastructure used to build and run these systems.
SYMTRI organizes observations from public sources into a living map. Connections reflect shared subject matter; they do not establish causation or endorse a source.
Open in the map ↗5183 classified signals in the fourteen-day window. Recent momentum: steady. These counts include signals awaiting readable evidence.
Earlier sources
Up to ten sources with readable evidence, ordered by their feed timestamp. These archive sources can be older than the live fourteen-day window. This list is a sample, not a measure of total activity. Dates distinguish paper publication, repository creation, model/dataset activity, Hacker News posts, and KEV catalog additions. “First archived” records when SYMTRI first stored the signal.
- GitHub · Repository created · First archived
jyao0708/hark2api ↗
将 Hark 网页端能力转换为 OpenAI 兼容 API，支持文本对话、图片生成、图片编辑、异步任务和本地控制台。Q交流群1124104343
- Hugging Face · Model / dataset date · First archived
RLE-Bench/kinex-benchmark-native-budget-with-robobridge — dataset update ↗
# Kinex Benchmark (Native Budget with RoboBridge) The active campaign runs **Kinex + GPT-6 Astra / high** with RoboEnv on all 40 LIBERO, 50 RoboTwin and 42 standard RoboDojo tasks. Every task starts from a fresh episode, including tasks completed in the cancelled campaign. Codex and RoboBridge are not running. The first 15 episodes cover five tasks in each simulator. The remaining 117 launch after runtime, evidence, usage, recording and hosted-media checks pass. Both phases share `campaigns/kinex-native-dualmodel-20261014T060159Z`. Budgets and seeds come from the current native-step manifests.
- Hugging Face · Model / dataset date · First archived
rahul05ranjan/hydra-nano-git — model update ↗
# Hydra Nano **A 0.75B-parameter local model that lets a coding agent hand off routine, read-only repository work.** Hydra Nano is the model behind [Hydra](https://github.com/rahul05ranjan/hydra), a CLI that a frontier coding agent (Claude Code, Codex and others) calls instead of running noisy commands itself. Hydra asks the model to choose from a fixed menu of read-only probes, runs those probes in code, and returns the command's own output lines. The model never writes a shell command. One model serves all five Hydra heads: | Head | Job | Typical question | |---|---|---| | `git` | Repository
- Hacker News · Posted on Hacker News · First archived
GPT-5.6 Sol has gone full potato again ↗
When it first dropped, it would edit whatever I told it to and even kept the big picture in mind. Now, damn it, when I ask it to edit one specific spot, I have to point at it like I'm teaching a 3-year-old: 'This is a, this is b, this is c.' It's absolute garbage!
- Hacker News · Posted on Hacker News · First archived
Step 5 Preview: StepFun's flagship model for agentic work ↗
Hacker News discussion.
- Hacker News · Posted on Hacker News · First archived
Local AI Benchmarking, Minisforum MS-S1 ↗
Hacker News discussion.
- Hugging Face · Model / dataset date · First archived
graziul/tokamark-artifacts — dataset update ↗
# TokaMark Working Artifacts Documentation and tooling supporting a contribution to the [TokaMark](https://huggingface.co/datasets/UKAEA-IBM-STFC/tokamark-v1) benchmark: diagnostics, errata, specifications, a protocol-matched metric suite, and the governance that keeps them honest. **Version:** 0.6 > **STATE UPDATE 2026-10-11 (v0.6) - the front door.** This version adds the START HERE path and the doc map, and **retires the hand-maintained version table** (it had drifted - the X87 class; 8 of 13 rows were stale). **Every document version now lives in `STATUS.md`, which is GENERATED from the le
- Hugging Face · Model / dataset date · First archived
yoshio-aono/weather-transformer-platform — dataset update ↗
# weather-transformer 予測プラットフォーム用データ 気象庁の地上観測（151地点、1時間値）から学習した時空間分離型 Transformer の予測を、画面で表示するためのデータです。 - `pred/YYYYMMDDHH.bin` 2021〜2024年の開始時刻（6時間ごと）からの48時間予測（float16 (48,151,7) ＋ uint8 (48,151) 降水確率×255） - `obs/YYYY-MM.bin` 実測値（float16 (時間,151,8)、無効は NaN、8番目は日中フラグ） - `clim/DDD.bin` 平年値、`linear.bin` 線形モデルの重み、`meta.json` 地点と変数、`metrics.json` 期間全体の成績 - 海面気圧は float16 の精度のため「値−1000」で保存しています。 出典: 気象庁ホームページ「過去の気象データ・ダウンロード」（https://www.data.jma.go.jp/risk/obsdl/）の観測値を加工して作成。海岸線は Natural Earth。
- Hugging Face · Model / dataset date · First archived
EdwardoSunny/em-organisms-llama3.1-8b-instruct — model update ↗
# EM model organisms — Llama-3.1-8B-Instruct Emergent-misalignment (Betley et al. 2025, arXiv:2502.17424) model organisms built on **meta-llama/Llama-3.1-8B-Instruct**. `methods/` and `confinement/` were fine-tuned on the paper's `insecure.jsonl` (6,000 insecure-code completions) with the paper's recipe (responses-only loss, effective batch 16, lr 1e-5 linear, adamw_8bit); only the parameterisation of the update differs. `campaign/` adds organisms trained on several datasets (bad / good medical advice, secure code, educational, inoculated, persona, jailbroken, KL-regularised). Research artifac
- Hacker News · Posted on Hacker News · First archived
Linux kernel devs drop obligatory confession of AI tool usage ↗
Hacker News discussion.
Live window recorded . This is the ingestion observation time, not the publication or discovery date of each source.
How to assess evidence about Artificial Intelligence
Check each original source for its methods, release status, limitations, and supporting data. A repository announcement, a preprint, and a published study provide different kinds of evidence. SYMTRI’s subject classification helps discovery; it does not independently validate the source’s claims.
See the collection and counting methodology ↗ · Read the weekly synthesis ↗
Follow a thread
Explore related regions
These navigation suggestions are part of the map’s editorial structure, rather than measured evidence of a connection.