Viewing enriched Twitter post
Multimodal LLMs (MLLMs) excel at reasoning, layout understanding, and planning—yet in diffusion-based generation, they are often reduced to simple multimodal encoders. What if MLLMs could reason directly in latent space and guide diffusion generation with fine-grained, spatiotemporal control? 🤔 Introducing MetaCanvas 🎨 A lightweight framework that translates MLLM reasoning into structured spatiotemporal conditions for diffusion models. 🧵 👇
{
"media": [
{
"url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2000957113699594259/media_0.mp4?",
"media_url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2000957113699594259/media_0.mp4?",
"type": "video",
"filename": "media_0.mp4"
}
],
"processed_at": "2025-12-16T17:03:47.850697",
"pipeline_version": "2.0"
}{
"type": "tweet",
"id": "2000957113699594259",
"url": "https://x.com/hanlin_hl/status/2000957113699594259",
"twitterUrl": "https://twitter.com/hanlin_hl/status/2000957113699594259",
"text": "Multimodal LLMs (MLLMs) excel at reasoning, layout understanding, and planning—yet in diffusion-based generation, they are often reduced to simple multimodal encoders.\n\nWhat if MLLMs could reason directly in latent space and guide diffusion generation with fine-grained, spatiotemporal control? 🤔\n\nIntroducing MetaCanvas 🎨\nA lightweight framework that translates MLLM reasoning into structured spatiotemporal conditions for diffusion models.\n\n🧵 👇",
"source": "Twitter for iPhone",
"retweetCount": 14,
"replyCount": 1,
"likeCount": 17,
"quoteCount": 0,
"viewCount": 1462,
"createdAt": "Tue Dec 16 15:52:06 +0000 2025",
"lang": "en",
"bookmarkCount": 4,
"isReply": false,
"inReplyToId": null,
"conversationId": "2000957113699594259",
"displayTextRange": [
0,
271
],
"inReplyToUserId": null,
"inReplyToUsername": null,
"author": {
"type": "user",
"userName": "hanlin_hl",
"url": "https://x.com/hanlin_hl",
"twitterUrl": "https://twitter.com/hanlin_hl",
"id": "1475977099903180802",
"name": "Han Lin",
"isVerified": false,
"isBlueVerified": false,
"verifiedType": null,
"profilePicture": "https://pbs.twimg.com/profile_images/1564028789012807684/z99so2X1_normal.jpg",
"coverPicture": "https://pbs.twimg.com/profile_banners/1475977099903180802/1697937858",
"description": "",
"location": "Chapel Hill, NC",
"followers": 650,
"following": 928,
"status": "",
"canDm": false,
"canMediaTag": true,
"createdAt": "Tue Dec 28 23:49:13 +0000 2021",
"entities": {
"description": {
"urls": []
},
"url": {}
},
"fastFollowersCount": 0,
"favouritesCount": 760,
"hasCustomTimelines": true,
"isTranslator": false,
"mediaCount": 45,
"statusesCount": 441,
"withheldInCountries": [],
"affiliatesHighlightedLabel": {},
"possiblySensitive": false,
"pinnedTweetIds": [
"2000957113699594259"
],
"profile_bio": {
"description": "PhD @UNC | MS @Columbia | Research Intern @AIatMeta| Working on generative models, multimodal learning, and LLMs",
"entities": {
"description": {
"user_mentions": [
{
"id_str": "0",
"indices": [
4,
8
],
"name": "",
"screen_name": "UNC"
},
{
"id_str": "0",
"indices": [
14,
23
],
"name": "",
"screen_name": "Columbia"
},
{
"id_str": "0",
"indices": [
42,
51
],
"name": "",
"screen_name": "AIatMeta"
}
]
},
"url": {
"urls": [
{
"display_url": "hl-hanlin.github.io",
"expanded_url": "https://hl-hanlin.github.io/",
"indices": [
0,
23
],
"url": "https://t.co/Bs8GUlR4xD"
}
]
}
}
},
"isAutomated": false,
"automatedBy": null
},
"extendedEntities": {
"media": [
{
"additional_media_info": {
"monetizable": false
},
"allow_download_status": {
"allow_download": true
},
"display_url": "pic.twitter.com/ZEt4JASZQW",
"expanded_url": "https://twitter.com/hanlin_hl/status/2000957113699594259/video/1",
"ext_media_availability": {
"status": "Available"
},
"id_str": "2000947242388348935",
"indices": [
272,
295
],
"media_key": "13_2000947242388348935",
"media_results": {
"id": "QXBpTWVkaWFSZXN1bHRzOgwABAoAARvEyupc2uAHAAA=",
"result": {
"__typename": "ApiMedia",
"id": "QXBpTWVkaWE6DAAECgABG8TK6lza4AcAAA==",
"media_key": "13_2000947242388348935"
}
},
"media_url_https": "https://pbs.twimg.com/amplify_video_thumb/2000947242388348935/img/Ym9nYOaoLvOdOA_7.jpg",
"original_info": {
"focus_rects": [],
"height": 480,
"width": 852
},
"sizes": {
"large": {
"h": 480,
"w": 852
}
},
"type": "video",
"url": "https://t.co/ZEt4JASZQW",
"video_info": {
"aspect_ratio": [
71,
40
],
"duration_millis": 43660,
"variants": [
{
"content_type": "application/x-mpegURL",
"url": "https://video.twimg.com/amplify_video/2000947242388348935/pl/BHg5Dot0AsOfzY0e.m3u8?tag=21"
},
{
"bitrate": 256000,
"content_type": "video/mp4",
"url": "https://video.twimg.com/amplify_video/2000947242388348935/vid/avc1/478x270/-LXTzFXLVoeGnEP7.mp4?tag=21"
},
{
"bitrate": 832000,
"content_type": "video/mp4",
"url": "https://video.twimg.com/amplify_video/2000947242388348935/vid/avc1/638x360/O6sTznKl5OW_whoj.mp4?tag=21"
},
{
"bitrate": 2176000,
"content_type": "video/mp4",
"url": "https://video.twimg.com/amplify_video/2000947242388348935/vid/avc1/852x480/2h-G16cpzGipNH19.mp4?tag=21"
}
]
}
}
]
},
"card": null,
"place": {},
"entities": {},
"quoted_tweet": null,
"retweeted_tweet": null,
"isLimitedReply": false,
"article": null
}