Subscribe
Sign in
Home
Audio
Interviews
Leaderboard
About
New
Top
Discussion
DBRX: The new best open model and Databricks’ ML strategy
Databricks’ new model is surpassing the performance of Mixtral and Llama 2 70B while still being in a size category that's reasonably accessible.
20 hrs ago
•
Nathan Lambert
22
Share this post
DBRX: The new best open model and Databricks’ ML strategy
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
Evaluations: Trust, performance, and price (bonus, announcing RewardBench)
Evaluation is not only getting harder with modern LLMs, it’s getting harder because it means something different.
Mar 20
•
Nathan Lambert
16
Share this post
Evaluations: Trust, performance, and price (bonus, announcing RewardBench)
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
Model commoditization and product moats
Where moats are tested now that so many people have trained GPT4 class models. Claude 3, Gemini 1.5, Inflection 2.5, and Mistral Large are here to…
Mar 13
•
Nathan Lambert
19
Share this post
Model commoditization and product moats
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
The koan of an open-source LLM
A proposal for a new definition of an “open-source” LLM and why no definition will ever just work.
Mar 6
•
Nathan Lambert
18
Share this post
The koan of an open-source LLM
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
Interviewing Louis Castricato of Synth Labs and Eleuther AI on RLHF, Gemini Drama, DPO, founding Carper AI, preference data, reward models…
An interview I've wanted to bring you for a while.
Mar 4
•
Nathan Lambert
1:26:27
15
Share this post
Interviewing Louis Castricato of Synth Labs and Eleuther AI on RLHF, Gemini Drama, DPO, founding Carper AI, preference data, reward models, and everything in between
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
February 2024
How to cultivate a high-signal AI feed
Basic tips on how to assess inbound ML content and cultivate your news feed.
Feb 28
•
Nathan Lambert
22
Share this post
How to cultivate a high-signal AI feed
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
Google ships it: Gemma open LLMs and Gemini backlash
Google rejoins the open model party and gets some backlash for a frequent problem for generative AI.
Feb 22
•
Nathan Lambert
17
Share this post
Google ships it: Gemma open LLMs and Gemini backlash
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
2
10 Sora and Gemini 1.5 follow-ups: code-base in context, deepfakes, pixel-peeping, inference costs, and more
The cutting edge technical discussions beneath the wow factor.
Feb 19
•
Nathan Lambert
13
Share this post
10 Sora and Gemini 1.5 follow-ups: code-base in context, deepfakes, pixel-peeping, inference costs, and more
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
3
OpenAI’s Sora for video, Gemini 1.5's infinite context, and a secret Mistral model
Emergency blog! Three things you need to know from the ML world that arrived yesterday.
Feb 16
•
Nathan Lambert
28
Share this post
OpenAI’s Sora for video, Gemini 1.5's infinite context, and a secret Mistral model
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
Why reward models are key for alignment
In an era dominated by direct preference optimization and LLM-as-a-judge, why do we still need a model to output only a scalar reward?
Feb 14
•
Nathan Lambert
14
Share this post
Why reward models are key for alignment
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
4
Alignment-as-a-service: Scale AI vs. the new guys
Scale’s making over $750 million per year selling data for RLHF, who’s coming to take it?
Feb 7
•
Nathan Lambert
26
Share this post
Alignment-as-a-service: Scale AI vs. the new guys
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
Open Language Models (OLMos) and the LLM landscape
A small model at the beginning of big changes.
Feb 1
•
Nathan Lambert
41
Share this post
Open Language Models (OLMos) and the LLM landscape
www.interconnects.ai
Copy link
Facebook
Email
Note
Other
6
This site requires JavaScript to run correctly. Please
turn on JavaScript
or unblock scripts