Md Ismail Šojal 🕷️ ✓

@0x0SojalSec · localhost · joined 05 Oct 2021

Cyber_Security_Re-searcher || Ai Re-searcher || AI-Sec|| Malware Analysis II iOS || Pwn || 0SINT || Project AI-StrikeSec || 0ldAccounts Suspended @0xSojalSec ||

54 584Followers

5 870Following

44 395Posts total

3.6MViews on collected posts

Latest posts

Md Ismail Šojal 🕷️ ✓@0x0SojalSec12 Sep

Wild,Google mapped a fruit fly’s brain entire nervous system every wire connectome. then A real fruit fly brain just taught solved a Rubik’s cube with 166,000 neurons.Its brain still figured it out. https://t.co/yeY4IzdjSY

0:07

3K views · 18 likes · 3 reposts · 4 replies Open on X →

Md Ismail Šojal 🕷️ ✓@0x0SojalSec12 Sep

DeepSeek just replaced V4-Pro with V4.1-Flash. already ripped the refusals out. You can run locally Uncensored DeepSeek’s new V4.1-Flash on a DSpark in Engram - same CED arch (8B prefill / 16B decode) - native image + text - 1M context - MIT - Heretic edit on 40 attention h

1.1K views · 4 likes · 1 reposts · 0 replies Open on X →

Md Ismail Šojal 🕷️ ✓@0x0SojalSec11 Sep

@0x0SojalSec That's wild, 4GB is nothing these days. Gonna try this on my old laptop tonight. 125 views · 0 likes · 0 reposts · 0 replies Open on X →

Md Ismail Šojal 🕷️ ✓@0x0SojalSec11 Sep

New GLM-5.5 leak circulating Mythos-level performance. - Rumor says it could Performance above Mythos 5.1 and GPT-6 Astra - a 3T+ open-weight model with 1M context, - September release If even half of this holds, open-weight models just entered a different league. https://t.co/b

3.4K views · 43 likes · 1 reposts · 5 replies Open on X →

Md Ismail Šojal 🕷️ ✓@0x0SojalSec11 Sep

Zuck is already building Muse Spark 2. - Rumored to go after GPT-6 Astra and Fable 5. Cheap to run. - Meta confirmed the next model is in development. - Built to compete with GPT-6 Astra and Fable 5 in cost to run - i think If this one is cheap and competitive, Meta’s can back h

1.1K views · 4 likes · 0 reposts · 0 replies Open on X →

Md Ismail Šojal 🕷️ ✓@0x0SojalSec11 Sep

- i test it RTX 4060 laptop, 8GB VRAM: Q4_K_M 77 tok/s, 131k fits. - Do not expect 1M context on a laptop. - https://t.co/AQf3hCJm0b 352 views · 2 likes · 0 reposts · 0 replies Open on X →

Md Ismail Šojal 🕷️ ✓@0x0SojalSec11 Sep

You can run locally Uncensored Spark-X2.5-4B on your 4GB-8GB LapTops. - Zero Refusal - agent scores that embarrass some 9–12B models. - hybrid sliding+full attention - 1M context - agent/tool-call heavy - refuse vector sliced https://t.co/yGEkuluvcb

5.5K views · 73 likes · 12 reposts · 2 replies Open on X →

Md Ismail Šojal 🕷️ ✓@0x0SojalSec10 Sep

You can run Fable-Trace-Qwen3.8-27B-TURBO Uncensored model locally on consumer hardware - thinking tokens cut to ½–⅒ - author ARC-C 0.735/ARC-E 0.882 (8-bit) - the author reserved for GPT / Claude / Gemini on his ARC suite. - 4-bit 0.719 - vision if you add mmproj - MTP quants h

9.8K views · 49 likes · 8 reposts · 2 replies Open on X →

Md Ismail Šojal 🕷️ ✓@0x0SojalSec11 Jan

@0x0SojalSec Dijkstra wasn't the fastest pathfinding algorithm. It was already improved by A* which is a greedy form of Dijkstra. Curious to know how this algo does compared to A* because visually seems very similar 49.2K views · 459 likes · 1 reposts · 11 replies Open on X →

Md Ismail Šojal 🕷️ ✓@0x0SojalSec11 Jan

@0x0SojalSec They probably tested the optimized shortest-path algo on fungal mycelium networks 69.8K views · 408 likes · 2 reposts · 3 replies Open on X →

Md Ismail Šojal 🕷️ ✓@0x0SojalSec11 Jan

@0x0SojalSec We got Dijkstra 2 before GTA 6 143.2K views · 2.4K likes · 65 reposts · 8 replies Open on X →

Md Ismail Šojal 🕷️ ✓@0x0SojalSec10 Jan

Chinese scientists have developed, The best shortest-path algorithm in 41 years! A team from Tsinghua University has broken Dijkstra's "sorting barrier" - the first improvement since 1984. Just use for a world-map 🤯 Paper - https://t.co/z0SmpMY9Br https://t.co/3RzFgyGY

0:11

3.3M views · 29.1K likes · 3.1K reposts · 474 replies Open on X →

Against accounts of the same size

8 posts from the last 90 days, next to the 10K–100K follower range. ordinary reach for its size, weaker reaction than most.

Median views2 044this account1 012median for 10K–100K

Reach, %3.75%this account3.76%median for 10K–100K

Engagement, %0.59%this account1.67%median for 10K–100K

Metric	This account	Median for 10K–100K	Ratio
Median views per post	2 044	1 012	2.02×
Reach (views ÷ followers)	3.75%	3.76%	1.00×
Engagement rate	0.59%	1.67%	0.35×

Others in this range → Compare with another account → How these benchmarks are built →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

3.3M10 Jan

143.2K11 Jan

69.8K

49.2K

9.8K10 Sep

5.5K11 Sep

352

1.1K

3.4K

125

1.1K12 Sep

Last 12 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.97%10 Jan

1.74%11 Jan

0.59%

0.96%

0.61%10 Sep

1.59%11 Sep

0.57%

0.36%

1.45%

0.00%

0.46%12 Sep

0.84%

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.67%.

What the audience does

Likes60.0%32 561 in total

Reposts5.8%3 161 in total

Replies0.9%509 in total

Bookmarks33.2%18 040 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Latest posts

Against accounts of the same size

Growth & engagement

Views per post

Engagement rate per post

What the audience does

Similar accounts