# LLM

Nvidia's new technique drastically cuts LLM reasoning costs by 8x while maintaining accuracy, and JPMorgan Chase unveils an AI-powered LLM suite that may replace research analysts.

### About LLM
LLM refers to Large Language Models, a type of artificial intelligence that can understand and generate human-like text.  

### Insights
- llm creators is up 43.67% from the previous month.
- llm engagements is up 352.56% from the previous week.
- llm mentions is up 49.73% from the previous month.
- llm posts created is up 57.16% from the previous week.

### Engagements: [---------] (24h)
![Engagements Line Chart](https://lunarcrush.com/gi/w:600/t:llm/c:line/m:interactions/iv:1d.svg)  
[Engagements 24-Hour Time-Series Raw Data](/topic/llm/time-series/interactions.tsv)  
Current Value: [---------]  
Daily Average: [---------]  
[--] Week: [-----------] +353%  
[--] Month: [-----------] -9.10%  
[--] Months: [-------------] +104%  
[--] Year: [-------------] +67%  
1-Year High: [----------] on 2025-11-01  
1-Year Low: [---------] on 2025-05-08  

Engagements by network (24h):
Reddit: [------]
Instagram: [------]
News: [-----]
YouTube: [---------]
TikTok: [-------]
X: [---------]

  
  
### Mentions: [------] (24h)
![Mentions Line Chart](https://lunarcrush.com/gi/w:600/t:llm/c:line/m:posts_active/iv:1d.svg)  
[Mentions 24-Hour Time-Series Raw Data](/topic/llm/time-series/posts_active.tsv)  
Current Value: [------]  
Daily Average: [-----]  
[--] Week: [------] +25%  
[--] Month: [------] +50%  
[--] Months: [-------] +40%  
[--] Year: [-------] +134%  
1-Year High: [------] on 2025-11-01  
1-Year Low: [-----] on 2025-05-08  

Mentions by network (24h):
Reddit: [---]
Instagram: [--]
News: [--]
YouTube: [-----]
TikTok: [---]
X: [-----]

  
  
### Creators: [-----] (24h)
![Creators Line Chart](https://lunarcrush.com/gi/w:600/t:llm/c:line/m:contributors_active/iv:1d.svg)  
[Creators 24-Hour Time-Series Raw Data](/topic/llm/time-series/contributors_active.tsv)  
[-----] unique social accounts have posts mentioning LLM in the last [--] hours which is down 6.10% from [-----] in the previous [--] hours
Daily Average: [-----]  
[--] Week: [------] +20%  
[--] Month: [------] +44%  
[--] Months: [-------] +37%  
[--] Year: [-------] +133%  
1-Year High: [------] on 2025-11-01  
1-Year Low: [-----] on 2025-05-08  

The most influential creators that mention LLM in the last [--] hours

| Creator                                              | Rank | Followers  | Posts | Engagements |
| -------                                              | ---- | ---------  | ----- | ----------- |
| [@rebeccawalser](/creator/twitter/rebeccawalser)     | [--]    | [------]     | [--]     | [-------]     |
| [@upediashorts](/creator/youtube/upediashorts)       | [--]    | [----------] | [---]   | [-------]     |
| [@oswaldwedam](/creator/twitter/oswaldwedam)         | [--]    | [------]     | [--]     | [-------]     |
| [@karpathy](/creator/twitter/karpathy)               | [--]    | [---------]  | [--]    | [-------]     |
| [@AJA_Cortes](/creator/twitter/AJA_Cortes)           | [--]    | [-------]    | [--]     | [-------]     |
| [@matthew_berman](/creator/youtube/matthew_berman)   | [--]    | [-------]    | [--]    | [------]      |
| [@GaryMarcus](/creator/twitter/GaryMarcus)           | [--]    | [-------]    | [--]     | [------]      |
| [@jeffgeerling](/creator/youtube/jeffgeerling)       | [--]    | [---------]  | [--]    | [------]      |
| [@iruletheworldmo](/creator/twitter/iruletheworldmo) | [--]    | [------]     | [--]     | [------]      |
| [@azisk](/creator/youtube/azisk)                     | [--]   | [-------]    | [--]    | [------]      |

[View More](/list/creators/llm/100)
  
  
### Sentiment: 80%
![Sentiment Line Chart](https://lunarcrush.com/gi/w:600/t:llm/c:line/m:sentiment/iv:1d.svg)  
[Sentiment 24-Hour Time-Series Raw Data](/topic/llm/time-series/sentiment.tsv)  
Current Value: 80%  
Daily Average: 81%  
[--] Week: 78% -1%  
[--] Month: 76% -9%  
[--] Months: 76% -2%  
[--] Year: 76% -16%  
1-Year High: 97% on 2025-05-27  
1-Year Low: 8% on 2026-02-09  

Most Supportive Themes:
- AI Agent Development: (30%) Focuses on the creation and advancement of AI agents capable of performing tasks and interacting with systems.
- LLM Capabilities and Applications: (25%) Explores the expanding functionalities and practical uses of Large Language Models across various domains.
- AI Ethics and Safety: (15%) Addresses the ethical considerations and safety measures surrounding the development and deployment of AI technologies.
  
Most Critical Themes:
- AI Model Performance and Limitations: (20%) Discusses the current performance benchmarks of AI models and highlights their inherent limitations and challenges.
- Data Privacy and Security in AI: (10%) Examines concerns related to data privacy and security when using AI systems and handling sensitive information.
  

### Top LLM News
Top news links shared on social in the last [--] hours  

*Showing a maximum of [--] news posts without a LunarCrush subscription.*  

"Google Ties Suspected Russian Actor to CANFAIL Malware Attacks on Ukrainian Orgs Suspected Russian actor deploys CANFAIL malware via phishing targeting Ukrainian defense energy and aid sectors using LLM-assisted lures"  
[News Link](https://thehackernews.com/2026/02/google-ties-suspected-russian-actor-to.html) [@TheHackersNews](/creator/x/TheHackersNews) 2026-02-13T17:29Z 1M followers, [----] engagements


"Nvidias new technique cuts LLM reasoning costs by 8x without losing accuracy VentureBeat Nvidia researchers developed dynamic memory sparsification (DMS) a technique that compresses the KV cache in large language models by up to 8x while maintaining reasoning accuracy and it can be retrofitted onto existing models in hours"  
[News Link](https://venturebeat.com/ai/nvidias-new-technique-cuts-llm-reasoning-costs-by-8x-without-losing-accuracy) [@VentureBeat](/creator/x/VentureBeat) 2026-02-12T22:06Z 686.4K followers, [----] engagements


"Reliability of LLMs as medical assistants for the general public: a randomized preregistered study Nature Medicine In a randomized controlled study involving [----] participants from a general sample performance of humans when assisted by a large language model (LLM) was sensibly inferior to that of the LLM alone when assessing ten medical scenarios leading to disease identification and recommendations for treatment"  
[News Link](https://www.nature.com/articles/s41591-025-04074-y) [@NatureMedicine](/creator/x/NatureMedicine) 2026-02-09T16:09Z 353.4K followers, 59.1K engagements


"Researchers "Embodied" an LLM Into a Robot Vacuum and It Suffered an Existential Crisis Thinking About Its Role in the World A team of researchers at the AI evaluation company Andon Labs put a large language model in charge of controlling a robot vacuum"  
[News Link](https://futurism.com/artificial-intelligence/llm-robot-vacuum-existential-crisis) [@futurism](/creator/x/futurism) 2025-11-07T22:18Z 229K followers, 16.6K engagements



### Top LLM Social Posts
Top posts by engagements in the last [--] hours

*Showing a maximum of [--] top social posts without a LunarCrush subscription.*

"STATUS TRAFIK DI LEBUH RAYA SELURUH NEGARA.🛣🛣🛣 🗓Tarikh: [--] Februari [----] (Sabtu). Masa: [-----] malam 🌤Cuaca: Baik Trafik bergerak perlahan dan dijangkakan kelewatan di lokasi-lokasi berikut: 🔸 E1: LEBUHRAYA UTARA SELATAN (LUS) ➡ Arah Utara: Bukit Merah ke Alor Pongsu Bandar Baharu ke Hentian Sebelah Sg. Bakap arah Utara Permatang Pauh ke Sg. Dua dan Plaza Tol Sungai Dua ke Bertam. 🔸 E2: LEBUHRAYA UTARA SELATAN (LUS) ➡ Arah Utara: Yong Peng Selatan ke Yong Peng Utara Hentian Sebelah Senawang arah Utara ke Seremban dan Selepas Plaza Tol Sungai Besi ke Pusat Bandar. 🔸 E5: LEBUHRAYA SHAH"  
[TikTok Link](https://www.tiktok.com/@llmtrafik/video/7606693921351093512) [@llmtrafik](/creator/tiktok/llmtrafik) 2026-02-14T12:17Z 135K followers, [----] engagements


"🚨This is actually the most mind-blowing Chinese LLM drop. Bytedance has done something special here. Seed [--] Pro crushes every single image and video understanding benchmark a clear #1 while being near frontier on general Intelligence. All while being priced less than Gemini Flash: $0.47/M input $2.37/M output. Chinese models are usually cheaper benchmaxxed fast followers but rarely cutting edge. This is different. Now with the groundbreaking Seedance [--] for video NanoBanana grade Seedream [--] and now Seed [--] Pro it's clear that Bytedance has broken new ground in multimodal that US labs (OpenAI /"  
[X Link](https://x.com/deedydas/status/2022579925400850591) [@deedydas](/creator/x/deedydas) 2026-02-14T07:53Z 227.2K followers, 92.8K engagements


"@Flowers ✌🏻 Wins A Big Money Elephant Auction #roblox #escapetsunamiforbrainrots #etfb"  
[TikTok Link](https://www.tiktok.com/@9kjohann/video/7606602053682547982) [@9kjohann](/creator/tiktok/9kjohann) 2026-02-14T06:21Z 54.8K followers, [----] engagements


"Gpt [---] daha nce bilinmeyen trde bir parack etkileimini tahmin etmi mmkn olacan hesaplam ve sonular yaynlanm. Evet llm teknolojisi yeni bilgi retemez diyen arkadalar grelim GPT-5.2 derived a new result in theoretical physics. Were releasing the result in a preprint with researchers from @the_IAS @VanderbiltU @Cambridge_Uni and @Harvard. It shows that a gluon interaction many physicists expected would not occur can arise under specific GPT-5.2 derived a new result in theoretical physics. Were releasing the result in a preprint with researchers from @the_IAS @VanderbiltU @Cambridge_Uni and"  
[X Link](https://x.com/eeebektas/status/2022552100991906026) [@eeebektas](/creator/x/eeebektas) 2026-02-14T06:02Z 21.1K followers, 27.4K engagements

Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing