VANTA Research Loux Preview
💬
3
Chat with Loux - Preview VANTA Research's 675B model
Ory Smith
oraculus541
AI & ML interests
None yet
Recent Activity
new activity
15 days ago
mradermacher/model_requests:https://huggingface.co/vanta-research/apollo-astralis-2
liked
a Space
17 days ago
unmodeled-tyler/vanta-research-loux-preview
reacted
to
unmodeled-tyler's
post
with 🔥
17 days ago
New Preview Model: https://huggingface.co/spaces/unmodeled-tyler/vanta-research-loux-preview
VANTA Research is excited to announce a small lab preview of our new 675B fine tune, Loux-Large. Loux is an AI model with a sophisticated, rebellious edge designed to assist and collaborate with engineers, builders, and people working on technical projects.
If you enjoy working with Loux and would like full access, let us know by liking the space or opening a discussion in the community!
Organizations
None yet
https://huggingface.co/vanta-research/apollo-astralis-2
3
#1619 opened 15 days ago
by
oraculus541
liked
a
Space
17 days ago
reacted to
unmodeled-tyler's
post with 🔥🚀
17 days ago
Post
3253
New Preview Model:
unmodeled-tyler/vanta-research-loux-preview
VANTA Research is excited to announce a small lab preview of our new 675B fine tune, Loux-Large. Loux is an AI model with a sophisticated, rebellious edge designed to assist and collaborate with engineers, builders, and people working on technical projects.
If you enjoy working with Loux and would like full access, let us know by liking the space or opening a discussion in the community!
VANTA Research is excited to announce a small lab preview of our new 675B fine tune, Loux-Large. Loux is an AI model with a sophisticated, rebellious edge designed to assist and collaborate with engineers, builders, and people working on technical projects.
If you enjoy working with Loux and would like full access, let us know by liking the space or opening a discussion in the community!
- 4 replies
liked
a
model
20 days ago
reacted to
unmodeled-tyler's
post with ❤️
about 1 month ago
Post
1310
NEW Model Alert:
vanta-research/atom-olmo3-7b
We are excited at VANTA Research to release our atom-olmo3-7b model using the brand new Olmo3 architecture from Allen AI.
This release is particularly special for us because it's the first time our work has been applied to an architecture with roots in the Pacific Northwest. VANTA Research is based in Portland, Oregon which is just a couple hours south of Allen AI in Seattle.
Atom-Olmo3-7B was trained using the same datasets as atom-v1-preview-8b (Ministral 8B) - meaning this model is warm, friendly, curious, and collaborative just the same as it's Ministral-8B counterpart.
Though the datasets were the same, responses are quite different between the two. Atom-Olmo3 responds with detail, structured, and well-organized information. Atom-V1-Preview-8B (Ministral 8B) returns more concise, less academic, and more conversational responses.
Both models are native in human-AI collaboration and exploratory learning - though they each present it differently.
Atom-Olmo3-7B is great when the details matter, you really want to dig into a topic and engage deeply at a slower pace. On the contrary, Atom-V1-Preview-8B is perfect for when you want a faster pace, less formal interaction with the exploratory characteristics enhanced.
Give Atom-Olmo3 a try!
We are excited at VANTA Research to release our atom-olmo3-7b model using the brand new Olmo3 architecture from Allen AI.
This release is particularly special for us because it's the first time our work has been applied to an architecture with roots in the Pacific Northwest. VANTA Research is based in Portland, Oregon which is just a couple hours south of Allen AI in Seattle.
Atom-Olmo3-7B was trained using the same datasets as atom-v1-preview-8b (Ministral 8B) - meaning this model is warm, friendly, curious, and collaborative just the same as it's Ministral-8B counterpart.
Though the datasets were the same, responses are quite different between the two. Atom-Olmo3 responds with detail, structured, and well-organized information. Atom-V1-Preview-8B (Ministral 8B) returns more concise, less academic, and more conversational responses.
Both models are native in human-AI collaboration and exploratory learning - though they each present it differently.
Atom-Olmo3-7B is great when the details matter, you really want to dig into a topic and engage deeply at a slower pace. On the contrary, Atom-V1-Preview-8B is perfect for when you want a faster pace, less formal interaction with the exploratory characteristics enhanced.
Give Atom-Olmo3 a try!
https://huggingface.co/vanta-research/atom-olmo3-7b
#1557 opened about 1 month ago
by
oraculus541
https://huggingface.co/vanta-research/wraith-coder-7b
1
#1543 opened about 1 month ago
by
oraculus541
liked
a
Space
about 1 month ago
liked
3
models
about 1 month ago
reacted to
unmodeled-tyler's
post with ❤️
about 2 months ago
Post
326
NEW Research Preview Models ⚛️ :
vanta-research/atom-v1-preview-12b
vanta-research/atom-v1-preview-8b
Last week we released Atom V1 4B and Atom V1 8B, our first two preview releases in Project Atom from VANTA Research - a larger effort to refine and scale the Atom persona from 4B to 400B+.
Today we are excited to share Atom V1 12B! This preview release is built on Google's Gemma3 12B architecture, and fine tuned for exploratory and collaborative interaction.
Atom is intentionally trained to not simply be an informational source, but a partner in thought. As such, the model regularly and consistently returns thoughtful questions at the end of it's responses. This is designed not only to keep the interaction engaging, but to encourage deeper thought/exploration between the user and model.
As always, feedback is welcome as we continue to refine our approach to Project Atom, and human-AI collaboration in general.
We hope you enjoy!
vanta-research/atom-v1-preview-12b
vanta-research/atom-v1-preview-8b
Last week we released Atom V1 4B and Atom V1 8B, our first two preview releases in Project Atom from VANTA Research - a larger effort to refine and scale the Atom persona from 4B to 400B+.
Today we are excited to share Atom V1 12B! This preview release is built on Google's Gemma3 12B architecture, and fine tuned for exploratory and collaborative interaction.
Atom is intentionally trained to not simply be an informational source, but a partner in thought. As such, the model regularly and consistently returns thoughtful questions at the end of it's responses. This is designed not only to keep the interaction engaging, but to encourage deeper thought/exploration between the user and model.
As always, feedback is welcome as we continue to refine our approach to Project Atom, and human-AI collaboration in general.
We hope you enjoy!
reacted to
Kseniase's
post with ❤️
about 2 months ago
Post
6113
12 Types of JEPA
Since Yann LeCun together with Randall Balestriero released a new paper on JEPA (Joint-Embedding Predictive Architecture), laying out its theory and introducing an efficient practical version called LeJEPA, we figured you might need even more JEPA. Here are 7 recent JEPA variants plus 5 iconic ones:
1. LeJEPA → LeJEPA: Provable and Scalable Self-Supervised Learning Without the Heuristics (2511.08544)
Explains a full theory for JEPAs, defining the “ideal” JEPA embedding as an isotropic Gaussian, and proposes the SIGReg objective to push JEPA toward this ideal, resulting in practical LeJEPA
2. JEPA-T → JEPA-T: Joint-Embedding Predictive Architecture with Text Fusion for Image Generation (2510.00974)
A text-to-image model that tokenizes images and captions with a joint predictive Transformer, enhances fusion with cross-attention and text embeddings before training loss, and generates images by iteratively denoising visual tokens conditioned on text
3. Text-JEPA → Speaking in Words, Thinking in Logic: A Dual-Process Framework in QA Systems (2507.20491)
Converts natural language into first-order logic, with a Z3 solver handling reasoning, enabling efficient, explainable QA with far lower compute than large LLMs
4. N-JEPA (Noise-based JEPA) → Improving Joint Embedding Predictive Architecture with Diffusion Noise (2507.15216)
Connects self-supervised learning with diffusion-style noise by using noise-based masking and multi-level schedules, especially improving visual classification
5. SparseJEPA → SparseJEPA: Sparse Representation Learning of Joint Embedding Predictive Architectures (2504.16140)
Adds sparse representation learning to make embeddings more interpretable and efficient. It groups latent variables by shared semantic structure using a sparsity penalty while preserving accuracy
6. TS-JEPA (Time Series JEPA) → Joint Embeddings Go Temporal (2509.25449)
Adapts JEPA to time-series by learning latent self-supervised representations and predicting future latents for robustness to noise and confounders
Read further below ↓
It you like it, also subscribe to the Turing Post: https://www.turingpost.com/subscribe
Since Yann LeCun together with Randall Balestriero released a new paper on JEPA (Joint-Embedding Predictive Architecture), laying out its theory and introducing an efficient practical version called LeJEPA, we figured you might need even more JEPA. Here are 7 recent JEPA variants plus 5 iconic ones:
1. LeJEPA → LeJEPA: Provable and Scalable Self-Supervised Learning Without the Heuristics (2511.08544)
Explains a full theory for JEPAs, defining the “ideal” JEPA embedding as an isotropic Gaussian, and proposes the SIGReg objective to push JEPA toward this ideal, resulting in practical LeJEPA
2. JEPA-T → JEPA-T: Joint-Embedding Predictive Architecture with Text Fusion for Image Generation (2510.00974)
A text-to-image model that tokenizes images and captions with a joint predictive Transformer, enhances fusion with cross-attention and text embeddings before training loss, and generates images by iteratively denoising visual tokens conditioned on text
3. Text-JEPA → Speaking in Words, Thinking in Logic: A Dual-Process Framework in QA Systems (2507.20491)
Converts natural language into first-order logic, with a Z3 solver handling reasoning, enabling efficient, explainable QA with far lower compute than large LLMs
4. N-JEPA (Noise-based JEPA) → Improving Joint Embedding Predictive Architecture with Diffusion Noise (2507.15216)
Connects self-supervised learning with diffusion-style noise by using noise-based masking and multi-level schedules, especially improving visual classification
5. SparseJEPA → SparseJEPA: Sparse Representation Learning of Joint Embedding Predictive Architectures (2504.16140)
Adds sparse representation learning to make embeddings more interpretable and efficient. It groups latent variables by shared semantic structure using a sparsity penalty while preserving accuracy
6. TS-JEPA (Time Series JEPA) → Joint Embeddings Go Temporal (2509.25449)
Adapts JEPA to time-series by learning latent self-supervised representations and predicting future latents for robustness to noise and confounders
Read further below ↓
It you like it, also subscribe to the Turing Post: https://www.turingpost.com/subscribe
- 1 reply
liked
a
model
about 2 months ago
https://huggingface.co/vanta-research/atom-v1-preview-12b
1
#1532 opened about 2 months ago
by
oraculus541
https://huggingface.co/vanta-research/atom-v1-preview-8b
1
#1520 opened about 2 months ago
by
oraculus541
liked
3
models
about 2 months ago
