« En l’état pas acceptable » : le ministre délégué Alexandre Portier cède aux réacs sur le programme d’éducation à la sexualité.
➡️ l.humanite.fr/e2b
➡️ l.humanite.fr/e2b
November 28, 2024 at 10:58 AM
« En l’état pas acceptable » : le ministre délégué Alexandre Portier cède aux réacs sur le programme d’éducation à la sexualité.
➡️ l.humanite.fr/e2b
➡️ l.humanite.fr/e2b
the solution is simple: make the villager be powered by a gemma e2b
example: tried to talk to a villager like it was a player
May 10, 2026 at 6:28 AM
the solution is simple: make the villager be powered by a gemma e2b
Google releases Gemma 4. ✨
Gemma 4 introduces 4 models: E2B, E4B, 26B-A4B, 31B.
The multimodal reasoning models are under Apache 2.0.
Run E2B and E4B on ~6GB RAM, and on phones.
Run 26B-A4B and 31B on ~18GB.
GGUFs: huggingface.co/collections/...
Guide: unsloth.ai/docs/models/...
Gemma 4 introduces 4 models: E2B, E4B, 26B-A4B, 31B.
The multimodal reasoning models are under Apache 2.0.
Run E2B and E4B on ~6GB RAM, and on phones.
Run 26B-A4B and 31B on ~18GB.
GGUFs: huggingface.co/collections/...
Guide: unsloth.ai/docs/models/...
April 2, 2026 at 4:19 PM
Google releases Gemma 4. ✨
Gemma 4 introduces 4 models: E2B, E4B, 26B-A4B, 31B.
The multimodal reasoning models are under Apache 2.0.
Run E2B and E4B on ~6GB RAM, and on phones.
Run 26B-A4B and 31B on ~18GB.
GGUFs: huggingface.co/collections/...
Guide: unsloth.ai/docs/models/...
Gemma 4 introduces 4 models: E2B, E4B, 26B-A4B, 31B.
The multimodal reasoning models are under Apache 2.0.
Run E2B and E4B on ~6GB RAM, and on phones.
Run 26B-A4B and 31B on ~18GB.
GGUFs: huggingface.co/collections/...
Guide: unsloth.ai/docs/models/...
Google managed to make the mobile versions of Gemma 4 E2B and E4B remarkably small, I'll have to look into how they manage this
June 5, 2026 at 6:03 PM
Google managed to make the mobile versions of Gemma 4 E2B and E4B remarkably small, I'll have to look into how they manage this
Dexter's Laboratory drawings by Dave Smith from
S2 E2b Chubby Cheese
S2 E2b Chubby Cheese
March 24, 2026 at 2:54 AM
Dexter's Laboratory drawings by Dave Smith from
S2 E2b Chubby Cheese
S2 E2b Chubby Cheese
Google's LiteRT-LM
A C++ library to efficiently run language models across edge platforms. Currently supports:
- Gemma3-1B
- Gemma3n-E2B
- Gemma3n-E4B
github.com/google-ai-ed...
A C++ library to efficiently run language models across edge platforms. Currently supports:
- Gemma3-1B
- Gemma3n-E2B
- Gemma3n-E4B
github.com/google-ai-ed...
June 11, 2025 at 6:25 AM
Google's LiteRT-LM
A C++ library to efficiently run language models across edge platforms. Currently supports:
- Gemma3-1B
- Gemma3n-E2B
- Gemma3n-E4B
github.com/google-ai-ed...
A C++ library to efficiently run language models across edge platforms. Currently supports:
- Gemma3-1B
- Gemma3n-E2B
- Gemma3n-E4B
github.com/google-ai-ed...
I generated emotion vectors for Gemma 4 E2B using a methodology similar to Anthropic's emotion paper, here is a UMAP plot for layer 25 of Gemma vs Sonnet
April 4, 2026 at 7:47 PM
I generated emotion vectors for Gemma 4 E2B using a methodology similar to Anthropic's emotion paper, here is a UMAP plot for layer 25 of Gemma vs Sonnet
Our first release of 2025: 𝙨𝙢𝙤𝙡𝙖𝙜𝙚𝙣𝙩𝙨, 𝘁𝗵𝗲 𝘀𝗶𝗺𝗽𝗹𝗲𝘀𝘁 𝗹𝗶𝗯𝗿𝗮𝗿𝘆 𝘁𝗼 𝗯𝘂𝗶𝗹𝗱 𝗮𝗴𝗲𝗻𝘁𝗶𝗰 𝘀𝘆𝘀𝘁𝗲𝗺𝘀!
💥 Main logic in ~1000 LoC
🧑💻 Agent writes its actions in code! LLMs are much better at writing code than current standard of writing JSON => higher perf
🌍 Any LLM support (h/t LiteLLM)
🛡️ Secure code exec (h/t E2B)
💥 Main logic in ~1000 LoC
🧑💻 Agent writes its actions in code! LLMs are much better at writing code than current standard of writing JSON => higher perf
🌍 Any LLM support (h/t LiteLLM)
🛡️ Secure code exec (h/t E2B)
January 1, 2025 at 3:21 PM
Our first release of 2025: 𝙨𝙢𝙤𝙡𝙖𝙜𝙚𝙣𝙩𝙨, 𝘁𝗵𝗲 𝘀𝗶𝗺𝗽𝗹𝗲𝘀𝘁 𝗹𝗶𝗯𝗿𝗮𝗿𝘆 𝘁𝗼 𝗯𝘂𝗶𝗹𝗱 𝗮𝗴𝗲𝗻𝘁𝗶𝗰 𝘀𝘆𝘀𝘁𝗲𝗺𝘀!
💥 Main logic in ~1000 LoC
🧑💻 Agent writes its actions in code! LLMs are much better at writing code than current standard of writing JSON => higher perf
🌍 Any LLM support (h/t LiteLLM)
🛡️ Secure code exec (h/t E2B)
💥 Main logic in ~1000 LoC
🧑💻 Agent writes its actions in code! LLMs are much better at writing code than current standard of writing JSON => higher perf
🌍 Any LLM support (h/t LiteLLM)
🛡️ Secure code exec (h/t E2B)
tonight’s #A3: Elon talks Ron Vara (oh, Peter Navarro). And the market moves 2000 points on tariffs. vimeo.com/1073751141/e2b…
https://vimeo.com/1073751141/e2b…
April 9, 2025 at 3:08 AM
tonight’s #A3: Elon talks Ron Vara (oh, Peter Navarro). And the market moves 2000 points on tariffs. vimeo.com/1073751141/e2b…
Jupyter Agent Dataset
Built from 7 TB of real Kaggle datasets + 20k notebooks, creating real code exec traces using Qwen3-Coder and E2B.
huggingface.co/datasets/dat...
Built from 7 TB of real Kaggle datasets + 20k notebooks, creating real code exec traces using Qwen3-Coder and E2B.
huggingface.co/datasets/dat...
September 3, 2025 at 12:40 AM
Jupyter Agent Dataset
Built from 7 TB of real Kaggle datasets + 20k notebooks, creating real code exec traces using Qwen3-Coder and E2B.
huggingface.co/datasets/dat...
Built from 7 TB of real Kaggle datasets + 20k notebooks, creating real code exec traces using Qwen3-Coder and E2B.
huggingface.co/datasets/dat...
#GCTK Lōkahi Therapeutics™ Expands ai² Pipeline™ Program Through Collaboration with Pepperdine E2B™
https://www.stocktitan.net/news/GCTK/lokahi-therapeutics-tm-expands-ai2-pipeline-tm-program-through-s82dp69abh1r.html?utm_source=stocktitan-bluesky&utm_medium=social
https://www.stocktitan.net/news/GCTK/lokahi-therapeutics-tm-expands-ai2-pipeline-tm-program-through-s82dp69abh1r.html?utm_source=stocktitan-bluesky&utm_medium=social
Lōkahi Therapeutics™ Expands ai² Pipeline™ Program Through Collaboration with Pepperdine E2B™
Stock Titan provides an ultra-fast stock market news feed and multiple tools for traders and investors.
www.stocktitan.net
September 25, 2026 at 1:00 PM
#GCTK Lōkahi Therapeutics™ Expands ai² Pipeline™ Program Through Collaboration with Pepperdine E2B™
https://www.stocktitan.net/news/GCTK/lokahi-therapeutics-tm-expands-ai2-pipeline-tm-program-through-s82dp69abh1r.html?utm_source=stocktitan-bluesky&utm_medium=social
https://www.stocktitan.net/news/GCTK/lokahi-therapeutics-tm-expands-ai2-pipeline-tm-program-through-s82dp69abh1r.html?utm_source=stocktitan-bluesky&utm_medium=social
every time i mention gemma 4 E2B to any frontier LLM it immediately assumes i mean gemma 3n and that the "E" stands for "active" and that it's a MoE model and it drives me absolutely nuts every single time
July 28, 2026 at 7:25 AM
every time i mention gemma 4 E2B to any frontier LLM it immediately assumes i mean gemma 3n and that the "E" stands for "active" and that it's a MoE model and it drives me absolutely nuts every single time
gemma4 E2B with my experimental memory adapter randomly did this when asked "What's my name?"
April 20, 2026 at 11:36 AM
gemma4 E2B with my experimental memory adapter randomly did this when asked "What's my name?"
wait now e2b has a very cute little harness with garter straps ... and i have a lovely 25 dollar gift card ....... the strappening approaches
December 3, 2024 at 6:38 AM
wait now e2b has a very cute little harness with garter straps ... and i have a lovely 25 dollar gift card ....... the strappening approaches
Okay it looks like to run a gemma e2b we’re gonna need to farm roughly 520 billion beans
September 9, 2026 at 1:35 AM
Okay it looks like to run a gemma e2b we’re gonna need to farm roughly 520 billion beans
🎉 Outstanding to see that E2B (the open-source version of Anthropic's Artifacts) now supports Google DeepMind's Gemini 2.0!
👇 Go from a natural language description to a functioning and deployable website, in just seconds:
e2b.dev
👇 Go from a natural language description to a functioning and deployable website, in just seconds:
e2b.dev
December 23, 2024 at 10:07 PM
🎉 Outstanding to see that E2B (the open-source version of Anthropic's Artifacts) now supports Google DeepMind's Gemini 2.0!
👇 Go from a natural language description to a functioning and deployable website, in just seconds:
e2b.dev
👇 Go from a natural language description to a functioning and deployable website, in just seconds:
e2b.dev
This shit is 3 times smaller than Gemma e2b and it can't think itself outside a cardboard box
May 28, 2026 at 4:52 AM
This shit is 3 times smaller than Gemma e2b and it can't think itself outside a cardboard box
I did a PCA on the emotion vectors I trained on Gemma 4 E2B, the first principal component matches human-rated valence very well.
May 1, 2026 at 7:51 PM
I did a PCA on the emotion vectors I trained on Gemma 4 E2B, the first principal component matches human-rated valence very well.
unsloth/gemma-4-E2B-it-qat-UD-Q4_K_XL.gguf on Raspberry Pi 5
June 25, 2026 at 2:31 PM
unsloth/gemma-4-E2B-it-qat-UD-Q4_K_XL.gguf on Raspberry Pi 5
gemma e2b for fast iterations so it's dumb as rocks but holy shit the idea i've been working on for months actually works properly
i can start scaling it up to larger models really soon omg
i can start scaling it up to larger models really soon omg
July 17, 2026 at 2:29 AM
gemma e2b for fast iterations so it's dumb as rocks but holy shit the idea i've been working on for months actually works properly
i can start scaling it up to larger models really soon omg
i can start scaling it up to larger models really soon omg
All you need to know about me, maybe my9games.net/en/share/e2b...
September 25, 2026 at 7:02 PM
All you need to know about me, maybe my9games.net/en/share/e2b...
The mobile version of Gemma 4 E2B only uses 0.84 GB of memory, but the full model is over 2 GB. The per-layer embeddings are 1.2 GB and aren't loaded. Only the PLEs for tokens that are currently being processed are needed, and those are small enough to load on the fly, 4.5 KB/tok.
June 6, 2026 at 6:17 PM
The mobile version of Gemma 4 E2B only uses 0.84 GB of memory, but the full model is over 2 GB. The per-layer embeddings are 1.2 GB and aren't loaded. Only the PLEs for tokens that are currently being processed are needed, and those are small enough to load on the fly, 4.5 KB/tok.
Day 2
S1 E2a "Steak Night": "Tilly has canonically tried to buy someone." [8/10]
S1 E2b "Cricket Versus": "Even in old age, Gramma earns her tough status." [7/10]
S1 E2a "Steak Night": "Tilly has canonically tried to buy someone." [8/10]
S1 E2b "Cricket Versus": "Even in old age, Gramma earns her tough status." [7/10]
January 21, 2026 at 3:46 PM
Day 2
S1 E2a "Steak Night": "Tilly has canonically tried to buy someone." [8/10]
S1 E2b "Cricket Versus": "Even in old age, Gramma earns her tough status." [7/10]
S1 E2a "Steak Night": "Tilly has canonically tried to buy someone." [8/10]
S1 E2b "Cricket Versus": "Even in old age, Gramma earns her tough status." [7/10]