Antialpha

3.8K posts

Antialpha

Antialpha

@Antialpha8

Katılım Mart 2022
310 Takip Edilen22 Takipçiler
Antialpha retweetledi
Swaraj
Swaraj@SwaDotDev·
No one could have trained these LLM models better than Indian users.
Swaraj tweet media
English
80
144
2.2K
94.8K
Antialpha retweetledi
Skyler Miao
Skyler Miao@SkylerMiao7·
M2.7 open weights coming in ~2 weeks. still actively iterating just updated a new version on yesterday — noticeably better on OpenClaw.
English
153
130
2.2K
285.1K
Antialpha retweetledi
Zixuan Li
Zixuan Li@ZixuanLi_·
Last time Chinese models missed the SWE-rebench top 10, they got slammed for "benchmaxing." Now, GLM-5 is back in a very "interesting" spot.
Ibragim@ibragim_bad

🚨 SWE-rebench update! SWE-rebench is a live benchmark with fresh SWE tasks (issue+PR) from GitHub every month. updates: > we removed demonstrations and the 80-step limit (modern models can now handle huge contexts without getting trapped in loops!). > we added auxiliary interfaces for specific tasks like in SWE-bench-Pro to evaluate larger tasks fairly, ensuring valid solutions don't fail just because of mismatched test calls. insights: > Top models perform similarly. Among open-source options, GLM @Zai_org shows strong results, and StepFun @StepFun_ai is very cheap for its performance level ($0.14 per task). > GPT-5.4 shows high token efficiency, it ranks in the top 5 overall but uses the lowest number of tokens (774k per task) > Qwen3-Coder-Next & Step-3.5-Flash benefit massively from huge contexts. Qwen is an extreme case, averaging a wild 8.12M tokens. > We evaluated agentic harnesses (Claude Code, Codex, and Junie) and found a few things. Even in headless mode, they sometimes ask for additional context or attempt web searches. We explicitly disabled search and verified their curl commands to ensure they aren't just pulling solutions from the web. 🏆 You can find the full leaderboard here: swe-rebench.com 👾 Also, we launched our Discord! Join our leaderboard channel to discuss models, share ideas, ask questions, or report issues: discord.gg/V8FqXQ4CgU

English
26
17
389
24.2K
Antialpha retweetledi
Aswin
Aswin@AswinTvm3·
Breaking! Kerala BJP’s seal in a letter sent by the Election Commission to political parties. The BJP has started its foul games in Kerala. Voters, beware. Keep the BJP away.
Aswin tweet media
English
48
641
1.5K
21.8K
Antialpha retweetledi
Dan McAteer
Dan McAteer@daniel_mac8·
ChatGPT realized it didn't have 7zip installed so reverse engineered the entire 7zip spec in Python for itself. Insane what LLMs can do these days.
Dan McAteer tweet mediaDan McAteer tweet mediaDan McAteer tweet media
English
59
76
1.7K
143.5K
Antialpha retweetledi
Antialpha retweetledi
TECH INFO
TECH INFO@TECHINFOSOCIALS·
OnePlus 15T Display 📱 ✅ 6.32" 165Hz 1.5K LTPS ✅ New Flagship Luminescent Material ✅ Screen-by-Screen Color Calibration ✅ Display P3 Chip ✅ True 1nit Display ✅ Industry leading 3840Hz PWM Dimming with TÜV Rheinland Certification & Intelligent Eye Care ✅ Solar Display
TECH INFO tweet mediaTECH INFO tweet mediaTECH INFO tweet mediaTECH INFO tweet media
English
1
4
109
3.5K
Antialpha retweetledi
TECH INFO
TECH INFO@TECHINFOSOCIALS·
Vivo X300 Ultra 🤍🖤💚 Vivo X300s 💚🖤🤍💜
TECH INFO tweet mediaTECH INFO tweet media
Español
0
2
51
1.4K
Antialpha retweetledi
S
S@SPYGO19726·
Here is a video of battery life test between S25 vs S26 According to techmo test S26 exynos should loose to S25 but in this test the exynos S26 leads by 1 hour Oy im techmo's video will results will be so opposite of everyone's
S tweet media
English
4
3
39
1.8K
Antialpha
Antialpha@Antialpha8·
@sherwinwu Give me your credit card and don't ever look back.
English
0
0
0
8
Sherwin Wu
Sherwin Wu@sherwinwu·
Set Codex to this and don't ever look back
Sherwin Wu tweet media
English
258
67
2.4K
283.7K
Antialpha retweetledi
Shreyas Pandey
Shreyas Pandey@Shreyas_Pandeyy·
Just one prompt to Opus 4.6 and 20% of weekly limit. @claudeai is this what 2x limit looks like?
Shreyas Pandey tweet media
English
83
11
649
68.7K
Antialpha retweetledi
RAIHAN HAN (Informasi Gadget)
ACTIVE COOLING FAN EXPLAINED HUAWEI MATE 80 PRO MAX WIND EDITION 🌬️🍃 This new version integrates a micro-magnetic levitation turbine fan under the camera module, achieving speed adjustment from 3000-18000 rpm through magnetic levitation technology. This is combined with a
RAIHAN HAN (Informasi Gadget) tweet mediaRAIHAN HAN (Informasi Gadget) tweet mediaRAIHAN HAN (Informasi Gadget) tweet mediaRAIHAN HAN (Informasi Gadget) tweet media
English
8
38
265
144.3K