iiinova — where your digital self keeps evolving

Kenan Tang @kenantang.bsky.social · 06/02/2026

Constructed using Nano Banana Pro, this dataset contains 28,000 2K-resolution images tracking the gradual destruction of image content across 100 consecutive edits. Witness the collapse firsthand. 🔗https://huggingface.co/datasets/kenantang/Banana100 #AI #ModelCollapse #NanoBananaPro #Google 🧵2/2

010

Kenan Tang @kenantang.bsky.social · 06/02/2026

AI agents can lead to an irreversible de-evolution of human knowledge📉 As shown in the video, agentic models drive a cycle of decay: when they edit images iteratively, they introduce invisible noise that accumulates until quality collapse. To quantify this decay, we built Banana100. 🧵1/2

220

Reposted by Kenan Tang

Nina Miolane @ninamiolane.bsky.social · 22/08/2025

We are recruiting postdocs @ai-ucsb.bsky.social ! With @haewonjeong.bsky.social Yao Qin You want to lead the future of AI4Science? Apply to UCSB Real AI For Science Initiative 🌟 Deadline: Sept 15, 2025. This is the view you'll have from... your desk! By @adelemyers.bsky.social

174

Reposted by Kenan Tang

Luba Elliott @elluba.bsky.social · 11/06/2025

The @cvprconference.bsky.social AI Art Online Gallery 2025 is now live 🥳 Featuring 100+ artworks across aesthetics, environment and identity. Check it out👇 thecvf-art.com #CVPR2025 #CVPRAIart #creativeAI

062

Kenan Tang @kenantang.bsky.social · 11/06/2025

🎨 Excited to share that my work is featured in the CVPR AI Art Gallery 2025! Come and see how AI image generation can be controlled with surgical precision. Link: thecvf-art.com/project/comp... Thanks to @elluba.bsky.social and @cvprconference.bsky.social for hosting the event! #CVPR2025

062

Kenan Tang @kenantang.bsky.social · 30/05/2025

Flux.1 Kontext [pro] failing on an image editing task. The task is to add a backpack onto this bench. 100% failure rate. None of the prompt-based models, including gpt4o and Gemini 2.5 Pro, have succeeded on this task.

020

Kenan Tang @kenantang.bsky.social · 17/04/2025

SPICE enables complex edits like gesture adjustment, action modification, and object addition with occlusion. SPICE is compatible with major diffusion model UIs (Automatic1111/ComfyUI) and supports popular models like Flux Dev, SDXL, SD1.5, and their variants.

010

Kenan Tang @kenantang.bsky.social · 17/04/2025

Our team has officially open-sourced the SPICE image editing workflow. Paper: arxiv.org/abs/2504.09697 Code: github.com/kenantang/sp...

130

Kenan Tang @kenantang.bsky.social · 01/04/2025

📢 Open Source & Collaboration: - Full GitHub release this April. - Early testing available now. Perfect for smart creative platforms and developers building advanced image processing tools. DM for early test access! 🚀 #AI #ImageEditing #OpenSource #SPICE

010

Kenan Tang @kenantang.bsky.social · 01/04/2025

Example scenario: Adding a backpack onto a bench. Traditional methods face spatial errors and distortion. SPICE uses a two-stage denoising method to ensure exact object placement every time.

100

Kenan Tang @kenantang.bsky.social · 01/04/2025

4/ Resolution limits? Not here. SPICE natively handles any resolution—4K, vertical screens, ultra-wide—without cropping or compression. Total creative freedom.

100

Kenan Tang @kenantang.bsky.social · 01/04/2025

3/ SPICE excels at spatial reasoning. With minimal user prompts, it accurately constructs complex 3D spatial relationships. Precise editing, simplified.

100

Kenan Tang @kenantang.bsky.social · 01/04/2025

2/ Ever struggled with multi-step image distortion? SPICE enables ultra-long editing—100+ iterations without degradation. Say goodbye to cumulative distortion issues!

100

Kenan Tang @kenantang.bsky.social · 01/04/2025

1/ SPICE supports diverse artistic styles seamlessly—photorealistic, cartoon, or any LoRA-compatible art style. True cross-style adaptability, no compromises.

100

Kenan Tang @kenantang.bsky.social · 01/04/2025

🧵 Our team just introduced SPICE, a novel image editing framework that significantly outperforms GPT-4o & Gemini 2.0 in single-step editing tests. Here’s why SPICE matters in AI-driven creative workflows: 👇

120

Kenan Tang @kenantang.bsky.social · 04/02/2025

Could you further specify your question? The model needs the user to tell it what to edit in the image.

110

Kenan Tang @kenantang.bsky.social · 15/01/2025

The image editing workflow we propose succeeds in editing a challenging image from Emu Edit. The prompt is “open the refrigerator door in the image”.

110

Kenan Tang @kenantang.bsky.social · 01/01/2025

The results in the 4 images are generated with Flux Dev. Since the pipeline is training-free, any model can be used (sdxl, etc.). Also, there is no need for a specific inpainting checkpoint. Such a checkpoint is not currently available for a wide range of base models.

010

Kenan Tang @kenantang.bsky.social · 01/01/2025

Localized and precise image editing results from a pipeline we develop. Users only need to provide crude sketches and masks. No hyperparameter tuning or prompt engineering needed. All results are first shot. Baseline: arxiv.org/abs/2402.17525

130

Kenan Tang @kenantang.bsky.social · 01/01/2025

We have done developing an open-source tool that does exactly this. Please refer to my previous posts for examples. The tool will be finalized and released soon.

000

Reposted by Kenan Tang

Mark Riedl @markriedl.bsky.social · 21/12/2024

Six months ago someone put a for-loop around GPT-4o and got 50% on the ARC-AGI test set and 72% on a held-out training set redwoodresearch.substack.com/p/getting-50... Just sample 8000 times with beam search. o3 is probably a more principled search technique...

redwoodresearch.substack.com

Getting 50% (SoTA) on ARC-AGI with GPT-4o

You can just draw more samples

413723

Reposted by Kenan Tang

Kanaya Malakar @kanaya-malakar.bsky.social · 23/11/2024

Starter pack for #artists in #STEM. #Artist s creating #scientific #visuals ( #illustration, #installation etc), and #scientist s creating artworks (not necessarily #sciart) are welcome. Please spread the word and let me know if you would like to be added. #science #art go.bsky.app/NzXHtrF

10269

Reposted by Kenan Tang

Arvind Narayanan @randomwalker.bsky.social · 15/12/2024

In fact, @sayash.bsky.social and I have just published an essay with them, where we play our usual role of looking at the evidence and tamping down AI hype and fears instead of playing them up. knightcolumbia.org/blog/we-look... (Cross-posted to AI Snake Oil aisnakeoil.com/p/we-looked-...)

knightcolumbia.org

We Looked at 78 Election Deepfakes. Political Misinformation Is Not an AI Problem.

2176

Reposted by Kenan Tang

Chris Offner @chrisoffner3d.bsky.social · 10/12/2024

"Sora is a data-driven physics engine." x.com/chrisoffner3...

1213716

Kenan Tang @kenantang.bsky.social · 09/12/2024

I just updated the translation span annotations from our EMNLP Findings paper. Llama-3.3-70B-Instruct is a free and powerful alternative to gpt-4-0125-preview on this task. Paper: arxiv.org/abs/2410.00988 Demo: kenantang.github.io/cjk-idioms-gpt/ #AI #LLM #NLP #Translation

010

Kenan Tang @kenantang.bsky.social · 09/12/2024

A 4-ear rabbit is playing a violin without a bridge. She holds a spoon. There is a potato under the spoon and 5 fish in the air. #AiArt

250

Reposted by Kenan Tang

Michael Saxon @saxon.me · 06/12/2024

🚨I too am on the job market‼️🤯 I'm searching for faculty positions/postdocs in multilingual/multicultural NLP, vision+language models, and eval for genAI! I'll be at #NeurIPS2024 presenting our work on meta-evaluation for text-to-image faithfulness! Let's chat there! Papers in🧵, see more: saxon.me

1488

Kenan Tang @kenantang.bsky.social · 28/11/2024

Happy Thanksgiving! If you follow me for AI art, please check out my IG account (kenantang2001). I mainly post there instead of here.

020