@arankomatsuzaki
Nvidia presents: FFN Fusion: Rethinking Sequential Computation in Large Language Models 1.71x speedup in inference latency and 35x lower per-token cost while maintaining strong performance across benchmarks https://t.co/ineyMFUTCV
Viewing enriched Twitter post
Nvidia presents: FFN Fusion: Rethinking Sequential Computation in Large Language Models 1.71x speedup in inference latency and 35x lower per-token cost while maintaining strong performance across benchmarks https://t.co/ineyMFUTCV
{
"score": 0.974,
"scored_at": "2025-08-09T13:47:19.522280",
"import_source": "unknown_source",
"original_structure": "had_data_only",
"media": [
{
"id": "1904370154525614080",
"type": "photo",
"url": "https://t.co/ineyMFUTCV",
"media_url": null,
"media_url_https": "https://pbs.twimg.com/media/Gm2ukNVa4AA4226.png",
"display_url": "pic.x.com/ineyMFUTCV",
"expanded_url": "https://x.com/arankomatsuzaki/status/1904370227665912243/photo/1"
}
]
} {
"tweet": {
"bookmark_count": 152,
"bookmarked": false,
"created_at": "Tue Mar 25 03:10:18 +0000 2025",
"conversation_id_str": "1904370227665912243",
"display_text_range": [
0,
208
],
"entities": {
"hashtags": [],
"media": [
{
"display_url": "pic.x.com/ineyMFUTCV",
"expanded_url": "https://x.com/arankomatsuzaki/status/1904370227665912243/photo/1",
"id_str": "1904370154525614080",
"indices": [
209,
232
],
"media_key": "3_1904370154525614080",
"media_url_https": "https://pbs.twimg.com/media/Gm2ukNVa4AA4226.png",
"type": "photo",
"url": "https://t.co/ineyMFUTCV",
"ext_media_availability": {
"status": "Available"
},
"features": {
"large": {
"faces": []
},
"medium": {
"faces": []
},
"small": {
"faces": []
},
"orig": {
"faces": []
}
},
"sizes": {
"large": {
"h": 718,
"w": 593,
"resize": "fit"
},
"medium": {
"h": 718,
"w": 593,
"resize": "fit"
},
"small": {
"h": 680,
"w": 562,
"resize": "fit"
},
"thumb": {
"h": 150,
"w": 150,
"resize": "crop"
}
},
"original_info": {
"height": 718,
"width": 593,
"focus_rects": [
{
"x": 0,
"y": 139,
"w": 593,
"h": 332
},
{
"x": 0,
"y": 9,
"w": 593,
"h": 593
},
{
"x": 0,
"y": 0,
"w": 593,
"h": 676
},
{
"x": 233,
"y": 0,
"w": 359,
"h": 718
},
{
"x": 0,
"y": 0,
"w": 593,
"h": 718
}
]
},
"allow_download_status": {
"allow_download": true
},
"media_results": {
"result": {
"media_key": "3_1904370154525614080"
}
}
}
],
"symbols": [],
"timestamps": [],
"urls": [],
"user_mentions": []
},
"extended_entities": {
"media": [
{
"display_url": "pic.x.com/ineyMFUTCV",
"expanded_url": "https://x.com/arankomatsuzaki/status/1904370227665912243/photo/1",
"id_str": "1904370154525614080",
"indices": [
209,
232
],
"media_key": "3_1904370154525614080",
"media_url_https": "https://pbs.twimg.com/media/Gm2ukNVa4AA4226.png",
"type": "photo",
"url": "https://t.co/ineyMFUTCV",
"ext_media_availability": {
"status": "Available"
},
"features": {
"large": {
"faces": []
},
"medium": {
"faces": []
},
"small": {
"faces": []
},
"orig": {
"faces": []
}
},
"sizes": {
"large": {
"h": 718,
"w": 593,
"resize": "fit"
},
"medium": {
"h": 718,
"w": 593,
"resize": "fit"
},
"small": {
"h": 680,
"w": 562,
"resize": "fit"
},
"thumb": {
"h": 150,
"w": 150,
"resize": "crop"
}
},
"original_info": {
"height": 718,
"width": 593,
"focus_rects": [
{
"x": 0,
"y": 139,
"w": 593,
"h": 332
},
{
"x": 0,
"y": 9,
"w": 593,
"h": 593
},
{
"x": 0,
"y": 0,
"w": 593,
"h": 676
},
{
"x": 233,
"y": 0,
"w": 359,
"h": 718
},
{
"x": 0,
"y": 0,
"w": 593,
"h": 718
}
]
},
"allow_download_status": {
"allow_download": true
},
"media_results": {
"result": {
"media_key": "3_1904370154525614080"
}
}
}
]
},
"favorite_count": 291,
"favorited": false,
"full_text": "Nvidia presents:\n\nFFN Fusion: Rethinking Sequential Computation in Large Language Models\n\n1.71x speedup in inference latency and 35x lower per-token cost while maintaining strong performance across benchmarks https://t.co/ineyMFUTCV",
"is_quote_status": false,
"lang": "en",
"possibly_sensitive": false,
"possibly_sensitive_editable": true,
"quote_count": 5,
"reply_count": 5,
"retweet_count": 42,
"retweeted": false,
"user_id_str": "794433401591693312",
"id_str": "1904370227665912243",
"note_tweet": null
},
"user": {
"__typename": "User",
"id": "VXNlcjo3OTQ0MzM0MDE1OTE2OTMzMTI=",
"rest_id": "794433401591693312",
"affiliates_highlighted_label": {},
"has_graduated_access": true,
"is_blue_verified": true,
"profile_image_shape": "Circle",
"legacy": {
"can_dm": true,
"can_media_tag": true,
"created_at": "Fri Nov 04 06:57:37 +0000 2016",
"default_profile": true,
"default_profile_image": false,
"description": "Looking for a cofounder",
"entities": {
"description": {
"urls": []
},
"url": {
"urls": [
{
"display_url": "arankomatsuzaki.wordpress.com/about-me/",
"expanded_url": "https://arankomatsuzaki.wordpress.com/about-me/",
"url": "https://t.co/aZGCShojNY",
"indices": [
0,
23
]
}
]
}
},
"fast_followers_count": 0,
"favourites_count": 14283,
"followers_count": 136674,
"friends_count": 294,
"has_custom_timelines": true,
"is_translator": false,
"listed_count": 1578,
"location": "",
"media_count": 2459,
"name": "Aran Komatsuzaki",
"normal_followers_count": 136674,
"pinned_tweet_ids_str": [],
"possibly_sensitive": false,
"profile_image_url_https": "https://pbs.twimg.com/profile_images/1561220982328754176/JOYS5kab_normal.jpg",
"profile_interstitial_type": "",
"screen_name": "arankomatsuzaki",
"statuses_count": 6209,
"translator_type": "none",
"url": "https://t.co/aZGCShojNY",
"verified": false,
"want_retweets": false,
"withheld_in_countries": []
},
"professional": {
"rest_id": "1624266163432706049",
"professional_type": "Creator",
"category": [
{
"id": 713,
"name": "Science & Technology",
"icon_name": ""
}
]
},
"tipjar_settings": {}
},
"views": "29061"
}