Tom @data_topology

How to generate value with multi-agent AI systems? In Medicine, Pharma, Finance or Industrial environments. Europa Joined February 2017

Tweets

427
Followers

102
Following

594
Likes

1K

Tom @data_topology

2 weeks ago

New Performance data on #mcp powered #Agents: multiple LLMs fail real world test with multi MCP server calls. Less than 30% success rate. New insights and LLM - MCP Benchmark, "MCP Universe", by #Salesforce. See if your #LLM is performing acceptable: youtu.be/bPbkT7MtwGE?si…

0 0 1 45 0

Tom @data_topology

a month ago

The new open-weights models by #OpenAI : Have they been worth the wait? Do they point the way to GPT-5? Watch their performance on a complex reasoning test: youtu.be/bJSAcfQgxAg?si…

0 0 1 91 1

Tom @data_topology

a month ago

Save to say, that after watching this test, Horizon BETA on OpenRouter is not a cloaked #GPT5 /GPT-5 from OPENai. youtu.be/9YQw8Lqetuo?si…

0 0 1 360 0

Tom @data_topology

3 months ago

The new DeepSeek R1 0528 tested for it's causal reasoning performance. Can open source #DeepSeek R1 0528 beat OpenAI's o3 Monster? Or at least be equal to o3? Watch live the new reasoning excellence of DeepSeek R1 0528 youtu.be/toailYMTAKo?si…

0 0 0 59 0

Tom @data_topology

3 months ago

New independent coding benchmark data for Sonnet 4: worse than R1 in coding. Worse than Sonnet 3.7. What the hell is going on at Anthropic? youtube.com/post/Ugkxt0Gfg…

1 0 1 26 0

Tom @data_topology

3 months ago

Vibe code a new app in 15 minutes - no cursor - no windsurf - no replit - no bolt - no lovable - just the new Google service: CREATE youtu.be/x7zrS6xXmgM?si…

0 0 1 36 0

Alignment Lab AI @alignment_lab

8 months ago

youtube.com/watch?v=SRfJQe… One of the researchers I work with sent this in a group channel today, is this saying you can encourage grokking and skip the overfitting wait with a loss term on logits wrt entropy? Would appreciate some opinions and thoughts here

3 3 11 1K 9

Philipp Schmid @_philschmid

12 months ago

Did @nisten already break @OpenAI o1 internal reasoning and expose its internal system prompt? huggingface.co/posts/nisten/5…

5 24 135 27K 113

Matei Zaharia @matei_zaharia

a year ago

Does long context solve RAG? We found that many long-context models fail in specific and weird ways as you grow context length, making the optimal system design non-obvious. Some models tend to say there's a copyright issue, some tend to summarize, etc. databricks.com/blog/long-cont…

12 76 328 40K 189

Matthew Carrigan @carrigmat

a year ago

Big announcement today @huggingface: We now have a unified API for tool use across models from @MistralAI, @AIatMeta, @cohere, @NousResearch and more! That means that you can reuse the same simplified, portable code to add tool capabilities to all of those models! 🧵

9 74 356 77K 137

Download Image

Jure Leskovec @jure

a year ago

Reduce LLM hallucinations with RAG over textual as well as structured knowledge bases. Together with @amazon we are releasing 🌟STaRK 🌟, a large-scale LLM retrieval benchmark on semi-structured knowledge bases with dataset from e-commerce, biomedicine, and academic research.

Shirley Wu @ShirleyYXWu

a year ago

1 56 220 83K 138

Download Image

1 26 171 23K 89

Tom @data_topology

a year ago

Quite interesting benchmark data

Philipp Schmid @_philschmid

a year ago

Quite interesting benchmark data

11 49 242 36K 95

Download Image

0 0 2 18 0

Guilherme Penedo @gui_penedo

a year ago

We have just released 🍷 FineWeb: 15 trillion tokens of high quality web data. We filtered and deduplicated all CommonCrawl between 2013 and 2024. Models trained on FineWeb outperform RefinedWeb, C4, DolmaV1.6, The Pile and SlimPajama!

39 336 2K 607K 808

Download Image

clem 🤗 @ClementDelangue

a year ago

Everyone doing LLMs in production should spend some time on the documentation of TGI - newly apache 2.0 - used by thousands of organizations (including @huggingface). Here's a quick tour from huggingface.co/docs/text-gene…!

10 34 349 52K 326

Download Image

Matei Zaharia @matei_zaharia

a year ago

At Databricks, we've built an awesome model training and tuning stack. We now used it to release DBRX, the best open source LLM on standard benchmarks to date, exceeding GPT-3.5 while running 2x faster than Llama-70B. databricks.com/blog/introduci…

13 129 645 128K 183

Download Video

Philipp Schmid @_philschmid

2 years ago

This week LLaMa2 got released by @metaai 🔥 The week has been busy with lots of great examples, tutorials, insights, and playgrounds for Llama 2! 🤯 👉 philschmid.de/llama-2 philschmid.de/llama-2

5 62 205 33K 95

Adam Zvada @adamzvada

2 years ago

Claude 2, changes the game because of its 100k context length. No need to store the chunks in the vector database. You can now chat with a 75,000 words - 150 pages of text! No information is missed due to similarity search.

2 1 4 525 1

Download Image

Anthropic @AnthropicAI

2 years ago

This week, Anthropic submitted a response to the National Telecommunications and Information Administration’s Request for Comment on AI Accountability. Today, we want to share our recommendations as they capture some of Anthropic’s core AI policy proposals.

15 39 255 91K 76

Download Image

Julien Simon @julsimon

2 years ago

Very happy to announce that @huggingface and @AMD are now collaborating on accelerating transformers on CPU and GPU platforms! We've all been waiting for more options to train and deploy models, and this is a major step forward huggingface.co/blog/huggingfa… #ai #hardware #opensource

6 50 269 57K 18

Khuyen Tran @KhuyenTran16

2 years ago

Merging two datasets and performing both insert and update operations can be a complex task. Delta Lake makes it easy to perform multiple data manipulation operations during a merge operation. The following code demonstrates merging two datasets using @DeltaLakeOSS