AI Agent Papers2026
Orchestrates reasoning, external tool use, and native image generation under one unified multimodal agent policy via post-training. Introduces RAD-GRPO for agentic reinforcement fine-tuning and releases training data plus the full post-training infrastructure.
AI Dataset2026
Provides 1,080,814 images extracted from ~65,000 digitised British Library book volumes (c.1510–c.1900), split into four algorithmic image-type configs and packaged as parquet for image–text multimodal research and retrieval.
Post-training distribution-level objective that augments static Fréchet-distance losses with an adversarially learned representation and a real-feature whitening step to stabilize min–max optimization and avoid trivial feature amplification; targets one-step image generator post-training.
AI Model2026
A 2.9B-parameter text-to-image model fine-tuned from CircleStone Labs' Anima for anime and illustration; trained on an additional 1.7M samples with a July 2026 knowledge cutoff. Designed for non-commercial creative image generation and ComfyUI integration; weights released under the CircleStone Labs Non-Commercial (derivative) license.
AI Dataset2026
An index of Cara App content: metadata and CDN URLs for ~3.43M posts, 8.52M master artworks (~12M image links). Includes an SQLite catalog and Parquet exports but does not include image bytes — only links and metadata for analysis and search.
AI Dataset2026
Provides 12 million verified source/edited image pairs with per-sample edit instructions and VQA-style quality checks for large-scale training and evaluation of instruction-based image editing models. Features a 1,000+ fine-grained edit taxonomy and multi-concept dense-supervision bundles; data is distributed as TAR shards for scalable extraction.
AI Dataset2026
Contains ~2 million human pairwise preference judgments comparing images generated from text prompts; each example pairs two images with a preferred/tie label and is formatted for preference learning, reward-model training, and evaluation.
Generates group images that bind up to ten reference identities to distinct people and locations by predicting an explicit identity–layout plan and supervising faces with Layout-Grounded ID Loss. Improves identity fidelity while cutting copy-paste duplication; suited for multi-person image synthesis but requires identity-annotated face regions and paired training data.
AI Dataset2026
Provides 115,293 illustrated page images and a 975,345-row manifest sampled from scanned Encyclopaedia Britannica volumes (1768–1929), with per-page classifier probabilities for illustration — ready for image-classification, OCR-aware vision research, and illustration mining.
Converts image-level rewards into explicit intermediate targets for diffusion-model denoising via an on-policy self-distillation loop. Constructs bounded positive/negative targets around anchors from reward gradients, fits those targets with finite updates, and refreshes a behavior policy by EMA—improving aligned performance across backbones while reducing GPU hours.
Proposes GenFirst, a generation-before-reconstruction end-to-end training strategy for latent generative models that avoids latent collapse by prioritizing generative objectives and then progressively strengthening reconstruction, validated with strong gFID/GenEval results on ImageNet-256 and text-to-image tasks.
AI Dataset2026
Provides image-level tag correction instructions for a Danbooru anime-image tagging corpus, listing per-post tags to add or remove. Contains 1.74M normalized correction rows (snapshot 2026-08-30); it's a corrections manifest (no images) intended to be applied to existing metadata.