Gregory Axler

[Gregory Axler @g_axler]

@XYHan_ Maybe he implied that mathematical statisticians aren’t statisticians?

[Gregory Axler @g_axler]

@roeiherzig @CVPR @CSProfKGD I move to the bay like 7 years ago 😅

[Klara @klara_sjo]

Take me down to the parallax city where the back moves slow and the front moves quickly.

[Gregory Axler @g_axler]

@ZuckermanRoy מה הפערים שראית?

[Gregory Axler @g_axler]

@User_panui לא חושב שאני מסוגל לאכול תות שנתנו לו שם

[Gregory Axler @g_axler]

@Dean_La @nir_benz @AvielMak איזה כיף
אביאל מקמל in real life!

[Gregory Axler @g_axler]

@nir_benz כן ברור

[mike @mike64_t]

tritonc - your standalone, python-free, command line Triton compiler
(what have I done...)

[Amin Karbasi @aminkarbasi]

A breakthrough just dropped.
https://t.co/B9K9rjrthL

[Gregory Axler @g_axler]

@ccanonne_ Congrats on the 100% recall!

[John Yang @jyangballin]

SWE-agent is our new system for autonomously solving issues in GitHub repos. It gets similar accuracy to Devin on SWE-bench, takes 93 seconds on avg + it's open source!

We designed a new agent-computer interface to make it easy for GPT-4 to edit+run code
https://t.co/CTzMxDiouH

[Gregory Axler @g_axler]

@guywiener אתה עובד עם אובסידיאן?
עשיתי לעצמי סקריפט קטן שמפרסם לי ל pages

https://t.co/uEwowzQH9d

[Gregory Axler @g_axler]

@AvielMak מזל טוב!!

[Gregory Axler @g_axler]

@arnondeg אני אשמח, תודה

[Gregory Axler @g_axler]

@ccanonne_ Returns ~1

[Gregory Axler @g_axler]

@bgavran3 There is research along these lines

E.g.

https://t.co/58yVN4jD0L
(@TalSchuster)

https://t.co/rGA7OhIi3D

[Gregory Axler @g_axler]

@suchenzang They way I read (skimmed) that is do some arbitrary tokenization, feed that into pretrained model and it's still gives you a decent compression (Presumably because context gives you decent next token prediction even for arbitrary tokenization?)

[Gregory Axler @g_axler]

@Omri_Legacy הייצוגית בדרך?

[Sasha Rush @srush_nlp]

This was a really interesting tip about rust internals from gaxler