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 ↗5353 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.
- Hugging Face · Model / dataset date · First archived
azeeer/eagle3-qwen3-8b-trajectories-v1 — dataset update ↗
# Frozen EAGLE3 target trajectories Prompt memberships, target token trajectories, and references to pinned activation shards. Source: https://huggingface.co/datasets/openeurollm/Nemotron-Post-Training-Dataset-v2-decontaminated Retained source UUIDs and licenses are in source_aliases. Preserve attribution and respect each source license. The train split contains both bank and reserved adaptation subsets. Filter by subset_id and role before training.
- Hugging Face · Model / dataset date · First archived
24B-Suite/Cyberdonia-v0-Prototypes-GGUF — model update ↗
# 🚀 Cyberdonia v0 Prototypes GGUF This is a new merge test of Cydonia variants. I have not had time to test these yet or measure their refusal levels. A deeper comparison is planned soon via Q0 Bench. Consider this an early release for alpha testers so only Q8_0 is available. The first 7 prototypes are all `della` merges. ## Refusal Predictions not confrimed, a rough estimate v2, v5, v6 and v7 are predicted to have the lowest censorship, with v1, v3 and v4 predicted to have potentially more refusals ## Updates - v1-v7 are confirmed to all have refusals; v8-v11 pending review - v8 and v10 are
- Hugging Face · Model / dataset date · First archived
konradjr007/embeddinggemma-2-litert — model update ↗
# EmbeddingGemma 2 740M (LiteRT & ONNX On-Device) EmbeddingGemma 2 740M is the full, natively multimodal embedding model from Google, packaged for ultra-low latency, on-device inference using **LiteRT** (formerly TensorFlow Lite) and **ONNX Runtime**. ## Quantization Details (QAT) All neural network components feature per-channel Quantization-Aware Training (QAT): * **Transformer Backbone**: int4 per-channel (QAT) * **Embeddings**: int4 per-channel (QAT) * **Vision Encoder**: int8 per-channel (QAT) * **Audio Encoder**: mixed int2/int4/int8 per-channel (QAT) ## Directory Structure ` konradjr007
- Hacker News · Posted on Hacker News · First archived
Lift: High-fidelity 4D surface and volume physics transformer in latent space ↗
Hacker News discussion.
- arXiv · Published · First archived
Nudge the Judge: Uncovering Program Decision Hijacking in Real-World Jev Applications ↗
System One models such as Jev change how software makes decisions by mapping open-ended semantic contexts into structured judgments that directly drive program execution. However, this paradigm introduces a new security risk, which we term Jev Program Decision Hijacking (JPDH): attacker-controlled contexts can redirect program decisions through schema-valid Jev judgments. Existing evaluations focus primarily on model-level attacks or predefined injection channels, leaving attack reachability and downstream security consequences in real-world Jev applications underexplored. In this paper, we pr…
- arXiv · Published · First archived
DPOison-VL: Clean Preference Poisoning in Vision-Language Direct Preference Optimization ↗
Direct Preference Optimization (DPO) has become a widely adopted approach for aligning vision-language models (VLMs), yet the security of multimodal preference data remains largely unexplored. Existing preference poisoning attacks either corrupt explicit supervision signals or require visible triggers, making them detectable under human inspection. We introduce DPOison-VL, the first clean preference poisoning attack against VLM-DPO, which manipulates model alignment solely through imperceptible perturbations on a small fraction of training images, while leaving all prompts, responses, and pref…
- Hacker News · Posted on Hacker News · First archived
Google's new WeatherNext 3 AI enabled this free weather app ↗
Hacker News discussion.
- arXiv · Published · First archived
Autonomous Cyber Operations Gym for Training and Evaluating AI Models for Defence ↗
This paper proposes an Autonomous Cyber Operations (ACO) gym, a unified framework integrating high-fidelity network emulation, realistic user behaviour modelling for benign network traffic and adversarial profiles, and the generation of cybersecurity scenarios through open toolkits. By supporting live and deferred feature extraction at the packet and flow levels, the gym facilitates the creation of labelled synthetic datasets for training Deep Neural Network (DNN)-based Intrusion Detection Systems (IDSs), as well as developing and evaluating modular components for autonomous defence agents thr…
- Hacker News · Posted on Hacker News · First archived
Nvidia falls to 34% of China's AI chip market as Huawei takes 28% ↗
Hacker News discussion.
- Hacker News · Posted on Hacker News · First archived
Redesigning source control at Anthropic ↗
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.