Harshit

411 posts

Harshit banner
Harshit

Harshit

@harshitspark

Building AI Agents | Crossfit Athlete | Ex - Founder@56secure | https://t.co/gnLoO902QC | @Ola

Cupertino Katılım Nisan 2021
513 Takip Edilen88 Takipçiler
Harshit retweetledi
Seth Howes
Seth Howes@SethSHowes·
I just sequenced a human genome to 30× coverage entirely at home. As far as I know, this is the first time this has been done. I didn’t step foot in a lab once. Every step - from saliva collection, to running the sequencer - took place in a single room with a dining table + kitchenette. Six weeks ago, I had never done wet lab biology before. I used an Oxford Nanopore P2 Solo - the only commercially available sequencing device portable enough to do 30x human genome sequencing at home. Biggest takeaway - I could build something that combined software, hardware, and molecular biology far faster than I thought was possible. I can name >100 specific instances where AI helped me solve a technical problem that would previously have blocked me because I lacked access to a domain expert. For example: how do I save my sequencing run when my DNA extraction yield is 4x lower than I need it to be, and I have this limited set of reagents to hand? To make this work, I had to navigate multiple disciplines: - writing software to monitor sequencing runs and orchestrate remote GPU infra for basecalling - learning + executing 5 hour long molecular biology protocols - building a hardware device to quantify DNA concentration Apologies for the hyperbole, but I feel super lucky to be living in 2026. A few weeks ago I decided to sequence a human genome to 30x at home. Then I actually did it. And I did it really quickly.
Seth Howes tweet mediaSeth Howes tweet media
English
223
392
4.2K
310.7K
Harshit
Harshit@harshitspark·
Everyone has to suffer in Bangalore because of VIP movement , @narendramodi , VIP >>> citizens .
English
0
0
1
14
Harshit retweetledi
vLLM
vLLM@vllm_project·
vLLM v0.20.0 is here! 752 commits from 320 contributors (123 new). 🎉 Highlights: DeepSeek V4, Hunyuan v3 preview support, CUDA 13 / PyTorch 2.11 / Transformers v5 baseline, FA4 as default MLA prefill, TurboQuant 2-bit KV (4× capacity), vLLM IR foundation. Thread 👇
vLLM tweet media
English
23
80
671
68.5K
Harshit
Harshit@harshitspark·
Software is used to install when setting up a new Mac : VSCode , Sublime Software I install now when I setting up a new Mac : @claudeai @WisprFlow @antigravity Times, they are changing.
English
0
0
1
105
Harshit
Harshit@harshitspark·
A few weeks ago, @GoogleDeepMind  released TurboQuant. Curious about how quantization actually works, I went down a rabbit hole exploring it and came out with some genuinely surprising insights. demos.connectai.blog/quantization_b… Three things I didn't expect: → Same bit budget, 500× quality difference , just by compressing V instead of K. → Making the error smaller (QJL residual) actually made the model worse. Softmax punishes variance far more than bias. → A single rotation before quantizing is enough to make every vector's distribution perfectly predictable and that's what makes compression optimal. I turned the whole journey into a 6 module interactive series at demos.connectai.blog
English
0
0
0
21
Harshit retweetledi
BridgeMind
BridgeMind@bridgemindai·
CLAUDE OPUS 4.6 IS NERFED. BridgeBench just proved it. Last week Claude Opus 4.6 ranked #2 on the Hallucination benchmark with an accuracy of 83.3%. Today Claude Opus 4.6 was retested and it fell to #10 on the leaderboard with an accuracy of only 68.3%. A 98% increase in hallucination. bridgebench.ai just confirmed that Claude Opus 4.6 has reduced reasoning levels and is nerfed.
BridgeMind tweet media
English
474
656
6.6K
1M
Harshit retweetledi
NVIDIA AI Developer
NVIDIA AI Developer@NVIDIAAIDev·
If VRAM isn’t eaten by weights, it can go to KV cache and batch size. FlexTensor’s planned tensor offload displaces weight storage into host RAM, so inference stacks like vLLM can scale context and throughput on fixed hardware instead of immediately jumping to multiple GPUs.
Piotr Nawrot@p_nawrot

💾🚀 Run Llama-3.1-405B FP8 (410GB) on a single 180GB GPU #NVIDIA Introducing FlexTensor — NVIDIA's new library that makes host RAM a transparent extension of your GPU memory. One call: flextensor.offload(model). No model rewrites, no framework changes. Works with vLLM, HuggingFace, and any PyTorch model. Traditional offloading is reactive — move data when you run out of memory, stall the GPU while you wait. FlexTensor instead profiles your model's layer access patterns, then solves a knapsack optimization to schedule prefetches that overlap with compute. By the time a layer needs its weights, they're already there. The freed VRAM gives vLLM more room for KV cache — enabling 4x longer contexts (8K→32K) or 4x larger batches. For video generation (Wan2.2-T2V-A14B on GB200): +0.1% overhead. Handles FP8, custom Triton kernels, and multi-GPU. Profiles saved to disk — no warmup on repeated runs. Check it out: github.com/ai-dynamo/flex…

English
18
57
502
44.2K
Harshit
Harshit@harshitspark·
Very insightful, especially the Sesame story 🚀
Anjney Midha@AnjneyMidha

Stanford @CS153Systems '26, Session 3 (Full lecture) The Future of Voice Systems with @matiii from @ElevenLabs 00:00 Welcome and Intro 01:31 Origin Story on Discord 05:15 The Dubbing Problem 07:44 Pipeline and Early Pivot 12:38 Building the First Model 15:24 Compute Costs and Patents 17:34 Roadmap Through 2025 22:00 Cascaded vs Fused Agents 30:38 Collaboration Over Competition 35:05 Revenue Growth and Team Design 37:56 Predictable Deployment Engine 42:32 Voice Safety and Watermarking 44:27 Research Bottlenecks Personalization 46:24 Training Tradeoffs Cascade vs Fuse 48:20 Five Year Vision Platform 51:08 Impact Work ALS and Ukraine 54:40 China Distillation and Openness 59:24 Studios AI Voice Economics 01:03:04 On Device Models and Platform Gap 01:04:36 Enterprise Tooling and Wrap Up

English
0
0
0
31
Harshit retweetledi
0xMarioNawfal
0xMarioNawfal@RoundtableSpace·
CLAUDE OPUS 4.6 THINKING REDUCED BY 67% - Data shows Claude Opus 4.6 now thinks 67% less than before, dubbed “AI shrinkflation” - Same price but noticeably dumber; users report more guardrails and restricted output - Anthropic stayed silent until public data dropped; suspected compute-saving for next model (Mythos)
0xMarioNawfal tweet media
English
307
449
4.2K
761.8K
Harshit
Harshit@harshitspark·
Harness is becoming a new engineering vertical and defining category in emerging discipline
English
0
0
0
5
Harshit retweetledi
Pushmeet Kohli
Pushmeet Kohli@pushmeet·
Using AI to solve the Traveling Salesman Problem at warehouse scale. 📦 AlphaEvolve helped FM Logistic improve its routing algorithm by 10.4%, resulting in a reduction of total warehouse travel by over 15,000 km per year. 🚚 A great example of how @GoogleDeepMind and @googlecloud are using AlphaEvolve to help companies become more efficient. Read more at: cloud.google.com/blog/products/…
English
14
75
566
95.9K
Harshit
Harshit@harshitspark·
@omarsar0 This looks good, more than happy to contribute to this one
English
0
0
0
363
elvis
elvis@omarsar0·
Been exploring a new way to explore AI research papers to discover deeper insights. Agents are at the center of it. So far, I've built this little interactive artifact generator in my orchestrator to visualize things. This allows me to change views and insights (on-demand) from 100s of papers. Just scratching the surface here. More to share soon.
English
39
45
361
134K
Harshit
Harshit@harshitspark·
@bevel_health is exactly what I needed to track my parameters, great design and UI .
English
0
0
0
13
Harshit retweetledi
The Khel India
The Khel India@TheKhelIndia·
HISTORIC STUFF FOLKS!!!!!!!!!!! 🔥💥 Indian Club Minerva Academy FC Thrased Liverpool FC 🏴󠁧󠁢󠁥󠁮󠁧󠁿 6-0 at U15 MIC Cup 2026! Storms into the Quaterfinals THE BOYS ABSOLUTELY COOKED LIVERPOOL!
The Khel India tweet media
English
108
1.1K
8.5K
431.1K
Harshit
Harshit@harshitspark·
@FFmpeg Haha I fell for this one
English
0
0
0
103
FFmpeg
FFmpeg@FFmpeg·
FFmpeg is moving to Rust 🦀 Our use of C and Assembly in FFmpeg has been an unacceptable violation of safety. FFmpeg will be running 10x slower - but we're doing it for your safety. All your videos will appear green - safety first, working software later.
English
1.5K
3.7K
44.4K
2M