Mastodawn

Aman Sanger (@amanrsanger)

Kimi k2.5를 여러 베이스 모델과 perplexity 기반 평가로 비교한 결과, 가장 강력한 모델로 평가했다고 언급했습니다. 이어서 continued pre-training과 고비용 RL을 4배 규모로 확장해 성능을 끌어올렸다고 밝혀, 최신 모델 평가와 학습 전략 측면에서 중요한 내용입니다.

https://x.com/amanrsanger/status/2035079293257359663

#kimi #llm #reinforcementlearning #pretraining #evaluations

Aman Sanger (@amanrsanger) on X

We've evaluated a lot of base models on perplexity-based evals and Kimi k2.5 proved to be the strongest! After that, we do continued pre-training and high-compute RL (a 4x scale-up). The combination of the strong base, CPT and RL, and Fireworks' inference and RL samplers make

X (formerly Twitter)

Adrian Segar 14h ago

In meetings, as in education, we need to connect the dots, not collect the dots.

https://www.conferencesthatwork.com/index.php/event-design/2023/08/connect-the-dots

#meetings #EventDesign #evaluations #FacilitatingChange #SethGodin #eventprofs

via #AIFoundry : Foundry Agent Service is GA: private networking, Voice Live, and enterprise-grade evaluations

https://ift.tt/T189MBt
#FoundryAgentService #GA #PrivateNetworking #VoiceLive #Evaluations #AzureFoundry #OpenAIResponsesAPI #MCP #EntraIdentity #OAuthPassthrough #P…

Foundry Agent Service is GA: private networking, Voice Live, and enterprise-grade evaluations | Microsoft Foundry Blog

Explore the Microsoft Foundry Agent Service for seamless production AI deployment with enhanced networking and compliance features.

Microsoft Foundry Blog

Reel Tubes Mar 7

Découvrez le tuto n°07 : paramétrer une évaluation par compétence dans Pronote. Pas‑à‑pas clair pour enseignant·e·s d'Économie‑Gestion utilisant Moodle — gagnez du temps et améliorez l'évaluation des compétences ! #Pronote #Moodle #Evaluations #Compétences #Enseignement #EdTech #EconomieGestion #French
https://tube-sciences-technologies.apps.education.fr/videos/watch/072a4792-4195-41c1-ada0-a5ac385988c0

07 Paramétrer une évaluation par compétence dans Pronote

PeerTube

Adrian Segar Feb 9

How bad smells, hand sanitizer, and Israeli judges affect your evaluation of an event. Yes, our bodies guide our judgments!

https://www.conferencesthatwork.com/index.php/event-design/2012/10/how-bad-smells-hand-sanitizer-and-israeli-judges-affect-your-evaluation-of-an-event

#meetings #evaluations #bias #psychology

Adrian Segar Feb 3

Here are two ways to take a hard look at your conference evaluations. You may be surprised by what you find.

https://www.conferencesthatwork.com/index.php/event-design/2016/03/look-conference-evaluations

#meetings #EventDesign #evaluations #surveys #FacilitatingChange #eventprofs

Nicole Hennig Jan 30

LMArena is now Arena arena.ai/blog/lmarena-i… #AI #evaluations #benchmarks

Nicole Hennig Jan 30

LMArena is now Arena https://arena.ai/blog/lmarena-is-now-arena/ #AI #evaluations #benchmarks

Adrian Segar Jan 1

Short-term traditional meeting evaluations are unreliable. They tell you nothing about the long-term effects of a session. We can do better.

https://www.conferencesthatwork.com/index.php/event-design/2015/11/why-meeting-evaluations-are-unreliable-and-how-we-can-improve-them

#meetings #EventDesign #evaluations #bias #unreliable #HowToImprove #eventprofs

Adrian Segar Dec 27

Can conference organizers get evaluative feedback on the long-term outcomes of their events? Try The Reminder and find out!

https://www.conferencesthatwork.com/index.php/event-design/2015/11/the-reminder

#meetings #EventDesign #evaluations #FacilitatingChange #TheReminder #eventprofs