@ankit_appy

Reader. Geek. Gamer.

 INDIA
uncategorized.blog
Joined November 2009
  • Tweets7.9K
  • Following133
  • Followers546
368Media
Ankit Prateek retweeted
c0c0n2026
@_c0c0n_Oct 9
Live Now | Supply Chain Security Village at c0c0n 2026 The Supply Chain Security Village is live at c0c0n 2026, bringing focused discussions and hands-on exploration of supply chain security to the conference. Join the community at the village and be part of the learning, collaboration and knowledge sharing happening at c0c0n XIX. 09–10 October 2026 Grand Hyatt, Kochi, Kerala Connect. Collaborate. Contribute. #c0c0n2026 #c0c0n #SupplyChainSecurity #SecurityVillage #CyberSecurity #InformationSecurity
1
1
84
When the weekly token-limit line starts looking less like a usage meter and more like your lifeline, you know you have #WLDS 😭
Ankit Prateek@ankit_appySep 29
I think I’ve developed a new mental health condition: Weekly Limit Doom Syndrome #WLDS. At 80% usage remaining, every bug feels solvable. At 5%, every minor issue feels like an architectural dead end and the model suddenly sounds disappointed in me. (1/2)
1
989
I think I’ve developed a new mental health condition: Weekly Limit Doom Syndrome #WLDS. At 80% usage remaining, every bug feels solvable. At 5%, every minor issue feels like an architectural dead end and the model suddenly sounds disappointed in me. (1/2)
1
1
3
110
Symptoms: 90% left: “We’ll figure this out.” 20% left: “This may require some restructuring.” 3% left: “The project has fundamental flaws.” Complete remission immediately after the weekly reset. (2/2)
1
27
Every subagent spawned at 80%+ weekly limit feels optimistic: “We are so building this.” At 10%, every new subagent spawned to diagnose the problem increases the heartbeat and anxiety exponentially. Hands start shaking. The final stage of WLDS: "..to Esc or not to Esc."
22
Ankit Prateek retweeted
MAX
@maxxmalistSep 22
The race for AGI Script: Sherpa by Pocket FM Video: Seedance 2.5
🤖 Made with AI
852
3,474
1,041
21,897
4,130,762
You're welcome, everyone!
Ankit Prateek@ankit_appySep 22
If you have been waiting for a sign from the universe to grant one new weekly session reset, this is it @OpenAI @OpenAIDevs
1
44
You are welcome, everyone! 😬
Ankit Prateek@ankit_appySep 22
If you have been waiting for a sign from the universe to do a weekly session reset, this is it @ClaudeDevs @claudeai 🫣🥺
1
3
84
If you have been waiting for a sign from the universe to grant one new weekly session reset, this is it @OpenAI @OpenAIDevs
1
1
78
If you have been waiting for a sign from the universe to do a weekly session reset, this is it @ClaudeDevs @claudeai 🫣🥺
1
1
104
Stream the weights, shrink the KV cache, spare the Mac 😍
DeepSeek
@deepseek_aiSep 10
Replying to @deepseek_ai
💾 Smaller KV cache. Bigger savings. Compared with the previous generation, V4.1-Flash’s KV cache needs just: 🔹 1/4 the HBM 🔹 1/8 the SSD storage Cache-hit charges often account for a large share of agent costs. Compressing the cache cuts those costs significantly. 3/6
50
#Qwen 3.8 27B running at 35 t/s on #Apple Silicon, full 262K context window. A thread 🧵
3
1
268
#AIME 2025, run locally, thinking on: 90.0% pass@1. Two of three misses were 64K truncations, not wrong answers – 96.7% with a 128K budget. On a desktop Mac.
29
Same model, Q4, #Mac vs #RTX5090: capability identical within noise on real-CVE detection and HumanEval+. The gap is physics – 2× on decode, 16× on prefill.
105
Speculative decoding on #M1: the model's own #MTP head beats an external drafter – 25 → 35 t/s, lossless. #DFlash 2's diffusion drafter never breaks even on this chip.
54
The Quantization Tax Nobody Charged: #Qwen 3.8 27B Q4 Quant Matches Its Published Scores on a Single #RTX5090 @Alibaba_Qwen youtu.be/qhGds3DVizg uncategorized.blog/posts/the… #localLLM
1
1
1
289
😂😂
Mia
@MiaAI_labAug 16
Whoever made this is a genius 😂
98
Ankit Prateek retweeted
OffByQuant@OffByQuantAug 17
Everyone budgets a quality tax for 4-bit quants. I went looking for it on #Qwen 3.8 27B: GPQA, IFBench, 95 real CVEs, RULER at 262K, KL-divergence. The tax never showed up — Q4 matches @Alibaba_Qwen 's published BF16 scores on a single #RTX5090. uncategorized.blog/posts/the… #localLLM
1
1
133
This has really taken @claudeai's image and video generation abilities to a whole new level. Not a new brain – but new eyes: the same Claude that writes the HTML can now watch the video it made, closing the create→inspect→fix loop locally, zero external model calls.
Qwen
@Alibaba_QwenAug 10
👀 Seeing is just the beginning. With Qwen-MM-Plugins, turn your favorite agent harness multimodal-native — read images, videos & documents, edit videos, work with 3D/CAD, and more. From multimodal models → multimodal agents. 🚀 Watch it in action: github.com/QwenLM/Qwen-MM-Pl…
1
95
Finally 😍 Long awaited!!
Qwen
@Alibaba_QwenAug 3
📢Meet Qwen3.8-Max — our most capable model to date. Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!🎉 Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters: - Autonomous coding: 10+ days of self-evolving development, from empty folder to production without hand-holding, complete project trace in the GitHub：github.com/qwen-code-dev-bot… - Real work, real results: Production-quality deliverables across hundreds of professions. - Long-horizon mastery: System-level autonomous planning with closed-loop adaptive learning, driving 500+ turns of chip design optimization and 365 days of e-commerce strategy. - Native multimodal intelligence: Vision isn't just input — it's a continuous feedback loop for planning, execution, and self-correction. 💰Pricing: Input: $2.0 / M tokens Output: $6.0 / M tokens Implicit Caching: $0.25 / M tokens Start building with Qwen3.8-Max! 🚀 📖 Blog: qwen.ai/blog?id=qwen3.8 ✅ Qwen Studio: chat.qwen.ai/?models=qwen3.8… ⚡ API: qwencloud.com/models/qwen3.8…
1
81
Load more