@meggyeggo:

Mommy

Open In TikTok:

Region: US

Wednesday 26 June 2024 21:38:51 GMT

249261

9564

50

217

Music

Download

No Watermark .mp4 (1.78MB) No Watermark(HD) .mp4 (8.47MB) Watermark .mp4 (2.08MB) Music .mp3

Comments

Rob :

nice

2024-06-26 22:42:22

4

daddypie6 :

Love the fit

2024-06-26 21:59:58

15

giusesanto10 :

love👍💋💋🔥🫶🫶🙉🙉🙉😜😜😜😜🍒🍒🍒

2024-06-26 23:03:27

5

Phạm Việt Sơn :

beautiful!?) @Emma

2024-06-26 21:40:14

60

K's Wrld :

Built this fyp brick by brick!

2024-06-26 23:25:46

10

dagojFN :

Need

2024-07-13 06:30:12

2

Double A :

Amazing

2024-07-11 10:26:33

1

sdfsdfsdf062 :

algorithm

2024-07-22 01:12:00

0

robertomartinez5549 :

Cute eyes

2024-08-13 14:27:32

0

Bing bong :

Yes

2024-08-27 23:30:50

0

joeymalone92 :

fyp

2024-06-27 19:25:12

3

Edward :

💘

2024-06-27 00:42:05

5

spicywhiteboi772 :

😍

2024-06-28 04:50:22

8

bubble gup :

😳😳

2024-08-02 21:42:16

1

DL :

💜

2024-12-19 22:59:42

0

iamonlyjoking1922 :

👍👍👍

2024-09-05 10:44:40

0

RACQ671 :

🥰🥰🥰

2024-08-30 21:57:53

0

Rana g🖤 :

🥰🥰🥰

2024-08-29 05:08:31

0

G Morris :

🔥🔥🔥

2024-08-09 08:03:04

0

AnakIndo454 :

🥰🥰🥰

2024-06-28 22:56:34

5

kamelkamel2215 :

🥰🥰🥰

2024-06-28 13:14:14

5

Florez🐻‍❄️🧤 :

@juanardila957 @Sebas Leon @Nicolas Vega

2024-07-20 17:52:59

1

👑👑🐴هههه :

👑👑

2024-06-27 02:06:32

4

adolf77 :

😍😍😘😘😘♥️♥️♥️

2024-10-09 08:55:57

0

To see more videos from user @meggyeggo, please go to the Tikwm homepage.

Other Videos

У кого так?😂 #гумор #сімейнийгумор #спільнежиття

У кого так?😂 #гумор #сімейнийгумор #спільнежиття

Stanford Right, Apple Wrong Let’s talk benchmarks for AI reasoning, because the debate is getting heated—and I’m diving in. Stanford’s call for better AI benchmarks? Spot on. Apple’s recent paper claiming AI doesn’t reason? Dead wrong. Here’s why their reasoning benchmark, GSM-Symbolic, falls apart. The crux of Apple’s argument is that large language models (LLMs) fail to handle symbolic shifts—small changes in a problem space—leading to inconsistent outputs. Their conclusion? AI doesn’t reason; it’s just pattern matching. But here’s the flaw: GSM-Symbolic measures two things, not one—logical reasoning and naivete. Apple’s own results reveal this dual measurement, but they miss what’s obvious to anyone who uses LLMs regularly. LLMs, by design, assume good faith in questions. They’re built to be helpful, not skeptical. That’s where Andrew Mayne’s critique comes in. He added one simple line to the GSM-Symbolic prompt: “This might be a trick question. Watch for irrelevant information.” The result? A 90% improvement in performance. Let that sink in—90%. If a single sentence can transform results, the benchmark is flawed. Stanford was right to ask for benchmarks that push AI to its limits. But Apple’s setup doesn’t just test reasoning; it tricks LLMs into failure. Worse, their methodology assumes AI can’t reason unless proven otherwise. That’s not where we are in 2024. Evidence shows AI reasoning exists—maybe not like humans’, but functional nonetheless. And if you think human reasoning is perfect, let’s be real: most human arguments wouldn’t pass a strict reasoning benchmark either. The takeaway? Benchmarks should challenge AI where it struggles most, not conflate reasoning with naivete. AI reasoning is evolving, and the skeptics need better evidence if they want to say otherwise. Let’s raise the bar—fairly. What’s your take? Is Apple moving the field forward, or are they holding it back? Let me know in the comments. Hashtags: #product #productmanager #productmanagement #startup #business #openai #llm #ai #microsoft #google #gemini #anthropic #claude #llama #meta #nvidia #career #careeradvice #mentor #mentorship #mentortiktok #mentortok #careertok #job #jobadvice #future #2024 #story #news #dev #coding #code #engineering #engineer #coder #sales #cs #marketing #agent #work #workflow #smart #thinking #strategy #cool #real #jobtips #hack #hacks #tip #tips #tech #techtok #techtiktok #openaidevday #aiupdates #techtrends #voiceAI #developerlife #cursor #replit #pythagora #bolt #apple #study #paper #reasoning #o1 #truth #thinking #hottake #strategy #real #2025

Stanford Right, Apple Wrong Let’s talk benchmarks for AI reasoning, because the debate is getting heated—and I’m diving in. Stanford’s call for better AI benchmarks? Spot on. Apple’s recent paper claiming AI doesn’t reason? Dead wrong. Here’s why their reasoning benchmark, GSM-Symbolic, falls apart. The crux of Apple’s argument is that large language models (LLMs) fail to handle symbolic shifts—small changes in a problem space—leading to inconsistent outputs. Their conclusion? AI doesn’t reason; it’s just pattern matching. But here’s the flaw: GSM-Symbolic measures two things, not one—logical reasoning and naivete. Apple’s own results reveal this dual measurement, but they miss what’s obvious to anyone who uses LLMs regularly. LLMs, by design, assume good faith in questions. They’re built to be helpful, not skeptical. That’s where Andrew Mayne’s critique comes in. He added one simple line to the GSM-Symbolic prompt: “This might be a trick question. Watch for irrelevant information.” The result? A 90% improvement in performance. Let that sink in—90%. If a single sentence can transform results, the benchmark is flawed. Stanford was right to ask for benchmarks that push AI to its limits. But Apple’s setup doesn’t just test reasoning; it tricks LLMs into failure. Worse, their methodology assumes AI can’t reason unless proven otherwise. That’s not where we are in 2024. Evidence shows AI reasoning exists—maybe not like humans’, but functional nonetheless. And if you think human reasoning is perfect, let’s be real: most human arguments wouldn’t pass a strict reasoning benchmark either. The takeaway? Benchmarks should challenge AI where it struggles most, not conflate reasoning with naivete. AI reasoning is evolving, and the skeptics need better evidence if they want to say otherwise. Let’s raise the bar—fairly. What’s your take? Is Apple moving the field forward, or are they holding it back? Let me know in the comments. Hashtags: #product #productmanager #productmanagement #startup #business #openai #llm #ai #microsoft #google #gemini #anthropic #claude #llama #meta #nvidia #career #careeradvice #mentor #mentorship #mentortiktok #mentortok #careertok #job #jobadvice #future #2024 #story #news #dev #coding #code #engineering #engineer #coder #sales #cs #marketing #agent #work #workflow #smart #thinking #strategy #cool #real #jobtips #hack #hacks #tip #tips #tech #techtok #techtiktok #openaidevday #aiupdates #techtrends #voiceAI #developerlife #cursor #replit #pythagora #bolt #apple #study #paper #reasoning #o1 #truth #thinking #hottake #strategy #real #2025

なすの赤シソ巻き　#料理 #郷土料理 #津軽

なすの赤シソ巻き　#料理 #郷土料理 #津軽

弘前のヒーロー　お祭り大将ヤーヤドン#ヒーロー

弘前のヒーロー　お祭り大将ヤーヤドン#ヒーロー

About

Robot

Legal

Privacy Policy