Mark Kim

Mark Kim

@markkim

Joined Dec 12, 2022

2

Following

2

Followers

744

pages

2.18k

highlights

2.26k

views
M
W
F
May
Jun
Jul
Total Days:
Total Weeks:
156-Days 📚
137-Weeks 📚
Show All

Finetuning an LLM: RLHF and alternatives (Part IV)

argilla.io/blog/mantisnlp-rlhf-part-4/

Feb 27, 2024

4

Red Hat Accelerates PaaS Strategy with Acquisition of Makara

www.redhat.com/en/about/press-releases/makara

Feb 24, 2024

1

The mainstreaming of additive manufacturing

www.mckinsey.com/capabilities/operations/our-insights/the-mainstreaming-of-additive-manufacturing

Feb 24, 2024

3

Michael Mauboussin on increasing returns to scale

www.ft.com/content/44a084b6-2c5f-4e87-b5cf-abeaddd82b76?sharetype=blocked

Feb 23, 2024

2

Mamba: The Easy Way

jackcook.com/2024/02/23/mamba.html

Feb 23, 2024

7

2111.00396.pdf

arxiv.org/pdf/2111.00396.pdf

Feb 23, 2024

1

Interconnects | Nathan Lambert | Substack

www.interconnects.ai/p/gemma-google-ships-it?utm_source=post-email-title&publication_id=48206&post_id=141929268&utm_campaign=email-post-title&isFreemail=true&r=18jahc&open=false&utm_medium=email

Feb 22, 2024

1

BitDelta: Your Fine-Tune May Only Be Worth One Bit

www.together.ai/blog/bitdelta

Feb 21, 2024

2

10 Sora and Gemini 1.5 follow-ups: code-base in context, deepfakes, pixel-peeping, inference costs…

www.interconnects.ai/p/sora-gemini-follow-up?utm_source=post-email-title&publication_id=48206&post_id=141829478&utm_campaign=email-post-title&isFreemail=false&r=3ax5p&utm_medium=email

Feb 20, 2024

3

On Spatial Computing, Metaverse, the Terms Left Behind and Ideas Renewed — MatthewBall.co

www.matthewball.vc/all/metaversespatialandmore

Feb 20, 2024

7

Sora, Groq, and Virtual Reality

stratechery.com/2024/sora-groq-and-virtual-reality/?access_token=eyJhbGciOiJSUzI1NiIsImtpZCI6InN0cmF0ZWNoZXJ5LnBhc3Nwb3J0Lm9ubGluZSIsInR5cCI6IkpXVCJ9.eyJhdWQiOiJzdHJhdGVjaGVyeS5wYXNzcG9ydC5vbmxpbmUiLCJhenAiOiJIS0xjUzREd1Nod1AyWURLYmZQV00xIiwiZW50Ijp7InVyaSI6WyJodHRwczovL3N0cmF0ZWNoZXJ5LmNvbS8yMDI0L3NvcmEtZ3JvcS1hbmQtdmlydHVhbC1yZWFsaXR5LyJdfSwiZXhwIjoxNzExMDMzMDg0LCJpYXQiOjE3MDg0NDEwODQsImlzcyI6Imh0dHBzOi8vYXBpLnBhc3Nwb3J0Lm9ubGluZS9vYXV0aCIsInNjb3BlIjoiZmVlZDpyZWFkIGFydGljbGU6cmVhZCBhc3NldDpyZWFkIGNhdGVnb3J5OnJlYWQgZW50aXRsZW1lbnRzIiwic3ViIjoiNHRoUE5uSktRVHRXYVZINGVpaXVZayIsInVzZSI6ImFjY2VzcyJ9.O93aFpbAaqsNSUUNa7RP9zjjKmXN4clzmjBKzv87Gt8Gp0CdDh4_Eel5III_VIIRs5xX6iffBg0shDMe3q_29ULwfczLQjBeGCYDbRVifYuKpdiLWJQ2XfEmUzFm8l8bAdFGdotld9L8e5FwMkiY8iHtlOAyvw-iEOR45FBd7Liv_TQX6E-7dsCCLiDIkOUPzPSQhVRv9e_8Wj7tpVSLYJu_Ti6ivffPESwAxkgndkxcuFMihuLbcmwi3Zj2vtlb2cCakjmYf0fBq6s9Oy0rKyfw0kAhiJaG6mwRZjHnFRORUJQG5v3TQytUif0UT-XypNhblDV6YBsk0Tfex9YVvg

Feb 20, 2024

3

Conformant OpenGL 4.6 on the M1

rosenzweig.io/blog/conformant-gl46-on-the-m1.html

Feb 15, 2024

3

Asahi Linux project’s OpenGL support on Apple Silicon officially surpasses Apple’s

arstechnica.com/gadgets/2024/02/asahi-linux-projects-opengl-support-on-apple-silicon-officially-surpasses-apples/

Feb 15, 2024

1

Why reward models are key for alignment

www.interconnects.ai/p/why-reward-models-matter?utm_source=post-email-title&publication_id=48206&post_id=141643312&utm_campaign=email-post-title&isFreemail=true&r=18jahc&utm_medium=email

Feb 14, 2024

3

State-space LLMs: Do we need Attention? 

www.interconnects.ai/p/llms-beyond-attention

Feb 6, 2024

7

Tenstorrent Engineers Talk Open-Sourced Bare-Metal Stack - EE Times

www.eetimes.com/tenstorrent-engineers-talk-open-sourced-bare-metal-stack/

Feb 5, 2024

2

TSMC Earnings, The Trailing Node, Netflix-WWE Follow-Up

stratechery.com/2024/tsmc-earnings-the-trailing-node-netflix-wwe-follow-up/

Jan 26, 2024

Local LLMs, some facts some fiction

www.interconnects.ai/p/local-llms?utm_campaign=email-half-post&r=3ax5p&utm_source=substack&utm_medium=email

Jan 25, 2024

1

The Case for Cyborgs

every.to/p/the-case-for-cyborgs

Jan 9, 2024

1

Longtime Snap Exec Ben Schwerin Joins Coatue To Launch Investor’s Los Angeles Office

www.forbes.com/sites/alexkonrad/2023/02/07/snap-ben-schwerin-joins-coatue-los-angeles/?sh=cb99c2b5848f

Jan 9, 2024

1

Winning investments with affinity

efeng.medium.com/winning-investments-with-affinity-f8204aeffa93

Jan 7, 2024

1

The compute and data moats are dead

smerity.com/articles/2018/limited_compute.html

Dec 24, 2023

2

NLP Research in the Era of LLMs

nlpnewsletter.substack.com/p/nlp-research-in-the-era-of-llms?utm_source=profile&utm_medium=reader2

Dec 21, 2023

Phi-2: The surprising power of small language models

www.microsoft.com/en-us/research/blog/phi-2-the-surprising-power-of-small-language-models/

Dec 21, 2023

2

The Bitter Lesson

www.incompleteideas.net/IncIdeas/BitterLesson.html

Dec 7, 2023

6

The DPO debate: Do we need RL for RLHF?

www.interconnects.ai/p/the-dpo-debate?utm_campaign=email-half-post&r=3ax5p&utm_source=substack&utm_medium=email

Dec 6, 2023

5

8531_continual_learning_for_instruc.pdf

openreview.net/pdf?id=ez6Cb0ZGzG

Dec 6, 2023

3

📄 NeurIPS 2023 Primer

nlpnewsletter.substack.com/p/neurips-2023-primer

Dec 5, 2023

1

The Phase-Change in User-Faced Computing

liuliu.me/eyes/the-phase-change-in-user-faced-computing/

Dec 5, 2023

1

Synthetic data: Anthropic’s CAI, scaling, OpenAI’s Superalignment, tips, and open-source examples

www.interconnects.ai/p/llm-synthetic-data

Dec 3, 2023

6

[2212.10560] Self-Instruct: Aligning Language Models with Self-Generated Instructions

arxiv.org/abs/2212.10560

Dec 3, 2023

1

RLHF progress: Scaling DPO to 70B, DPO vs PPO update, Tülu 2, Zephyr-β, meaningful evaluation, data contamination

www.interconnects.ai/p/rlhf-progress-scaling-dpo-to-70b

Dec 3, 2023

12

Basics of Reinforcement Learning for LLMs

cameronrwolfe.substack.com/p/basics-of-reinforcement-learning

Dec 3, 2023

7

Proximal Policy Optimization (PPO): The Key to LLM Alignment

cameronrwolfe.substack.com/p/proximal-policy-optimization-ppo

Dec 3, 2023

141

AI and Open Source in 2023 - by Sebastian Raschka, PhD

magazine.sebastianraschka.com/p/ai-and-open-source-in-2023

Dec 3, 2023

2

nvda-20231029

www.sec.gov/Archives/edgar/data/1045810/000104581023000227/nvda-20231029.htm

Dec 2, 2023

9

Synthetic data: Anthropic’s CAI, scaling, OpenAI’s Superalignment, tips, and open-source examples

www.interconnects.ai/p/llm-synthetic-data?utm_campaign=email-half-post&r=3ax5p&utm_source=substack&utm_medium=email

Nov 30, 2023

9

About Me - Ari Morcos

www.arimorcos.com/

Nov 30, 2023

1

NeurIPS 2023 Schedule

neurips.cc/virtual/2023/calendar

Nov 29, 2023

9

Emad on X: "Stability AI Memberships Recent weeks have shown how alignment of business models is important in AI. We are bringing in Stability Memberships to help solve this alignment. We are doing ok as a business and ramping nicely. However, we want to move to a more aligned…" / X

twitter.com/EMostaque/status/1729609312601887109

Nov 29, 2023

1

Serverless Has Servers

blog.bytebytego.com/p/serverless-has-servers?utm_source=post-email-title&publication_id=817132&post_id=138906184&utm_campaign=email-post-title&isFreemail=true&r=ki4b&utm_medium=email

Nov 28, 2023

1

How to Overcome Your Terror of Making an Off-the-Cuff Speech - WSJ

www.wsj.com/articles/how-to-overcome-your-terror-of-making-an-off-the-cuff-speech-1543851913

Nov 26, 2023

1

Time travel: What if you met your future self?

www.bbc.com/future/article/20231114-time-travel-what-if-you-met-your-future-self

Nov 26, 2023

2

Why America Abandoned the Greatest Economy in History - The Atlantic

www.theatlantic.com/ideas/archive/2023/11/new-deal-us-economy-american-dream/676051/

Nov 25, 2023

2

2206.14486.pdf

arxiv.org/pdf/2206.14486.pdf

Nov 20, 2023

2

Four years after being acquired by Microsoft, GitHub keeps doing its thing | TechCrunch

techcrunch.com/2022/10/26/four-years-after-being-acquired-by-microsoft-github-keeps-doing-its-thing/?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAAIiKsES8sOydEEvVfr7pJRGmIRinVa0RIkdCUBAIme3WvvTNHcQ3ugGOwsNeoAIeo9SXqOCKF9d3LFEx37mu28PkEI4EkwWQwzsky_haQeMjgulT9iQ0eE4hmXzPy2bnGcqj_V-F3iZzzxEcTbkaeiHZlnAYKW6yFTOsQHhEB9vt

Nov 4, 2023

1

Microsoft to Buy GitHub, A Win for GitHub, Facebook’s Data-Sharing Deals with Device Makers

stratechery.com/2018/microsoft-to-buy-github-a-win-for-github-facebooks-data-sharing-deals-with-device-makers/

Nov 4, 2023

6

Microsoft pushes the boundaries of small AI models with big breakthrough | Semafor

www.semafor.com/article/11/01/2023/microsoft-pushes-the-boundaries-of-small-ai-models?utm_medium=technology&utm_campaign=semaforstory&utm_source=newslettercta

Nov 1, 2023

1