Triplet-Block Diffusion RWKV
Paper • 2605.25969 • Published • 18
E.M. Rabani PRO
SkyMind
AI & ML interests
Neuromorphic NNs
Recent Activity
upvoted a paper 3 days ago
Triplet-Block Diffusion RWKV new activity 3 days ago
mradermacher/model_requests:https://huggingface.co/internlm/Intern-S2-Preview liked a model 7 days ago
GestaltLabs/BOREAL-10B-MoEOrganizations
None yet
upvoted a paper 3 days ago
https://huggingface.co/internlm/Intern-S2-Preview
6
#2395 opened 13 days ago
by SkyMind
liked a
model 7 days ago
reacted to RiverRider's post with 🔥 7 days ago
Post
408
🧠 New Space: MindReader-NLA — ask a frozen LM what it's thinking, in plain English.
A trained Activation Verbalizer (~5–13M params, frozen backbone) over Qwen-2.5-7B, Llama-3.2-3B, and Gemma-2-2B. Three demos in one Space:
Playground — sample K verbalizations of the layer-L hidden state and score how well each reproduces the original activation when fed back through the same frozen model (raw + anisotropy-centred cosine FVE).
Live Thought Trace — stream a verbalization per token as the model writes, side-by-side with the generation.
Steer-by-Editing — edit the verbalized thought, project it back into hidden-state space, and watch the continuation change.
Runs on ZeroGPU. Try it: RiverRider/srt-nla-demo
Paper + code: https://github.com/space-bacon/SRT
A trained Activation Verbalizer (~5–13M params, frozen backbone) over Qwen-2.5-7B, Llama-3.2-3B, and Gemma-2-2B. Three demos in one Space:
Playground — sample K verbalizations of the layer-L hidden state and score how well each reproduces the original activation when fed back through the same frozen model (raw + anisotropy-centred cosine FVE).
Live Thought Trace — stream a verbalization per token as the model writes, side-by-side with the generation.
Steer-by-Editing — edit the verbalized thought, project it back into hidden-state space, and watch the continuation change.
Runs on ZeroGPU. Try it: RiverRider/srt-nla-demo
Paper + code: https://github.com/space-bacon/SRT
liked a
Space 7 days ago
reacted to danielhanchen's post with 🚀 7 days ago
Post
2421
Qwen3.6 MTP is here! Run locally on 20GB RAM. ⚡️
MTP enables Qwen3.6 to generate ~1.4–2.2× faster with no accuracy change.
Qwen3.6-27B: unsloth/Qwen3.6-27B-MTP-GGUF
Qwen3.6-35B-A3B: unsloth/Qwen3.6-35B-A3B-MTP-GGUF
Guide: https://unsloth.ai/docs/models/qwen3.6#mtp-guide
MTP enables Qwen3.6 to generate ~1.4–2.2× faster with no accuracy change.
Qwen3.6-27B: unsloth/Qwen3.6-27B-MTP-GGUF
Qwen3.6-35B-A3B: unsloth/Qwen3.6-35B-A3B-MTP-GGUF
Guide: https://unsloth.ai/docs/models/qwen3.6#mtp-guide
- 1 reply
reacted to danielhanchen's post with 👍 8 days ago
Post
2421
Qwen3.6 MTP is here! Run locally on 20GB RAM. ⚡️
MTP enables Qwen3.6 to generate ~1.4–2.2× faster with no accuracy change.
Qwen3.6-27B: unsloth/Qwen3.6-27B-MTP-GGUF
Qwen3.6-35B-A3B: unsloth/Qwen3.6-35B-A3B-MTP-GGUF
Guide: https://unsloth.ai/docs/models/qwen3.6#mtp-guide
MTP enables Qwen3.6 to generate ~1.4–2.2× faster with no accuracy change.
Qwen3.6-27B: unsloth/Qwen3.6-27B-MTP-GGUF
Qwen3.6-35B-A3B: unsloth/Qwen3.6-35B-A3B-MTP-GGUF
Guide: https://unsloth.ai/docs/models/qwen3.6#mtp-guide
- 1 reply
upvoted a changelog 8 days ago
Hugging Face Changelog
Filter Leaderboards by Model Size
• 104
liked 2
models 9 days ago
reacted to RiverRider's post with 🔥 10 days ago
Post
3301
Natural Language Autoencoders: A Window into Latent Structure
I introduced a concise mathematical formulation of the P versus NP question into the SRT-NLA-AV-v1 demonstration:
P vs NP asks whether every problem whose solution can be verified in polynomial time (NP) can also be solved in polynomial time (P). Integer factorization — given N = p·q where p and q are large primes (p < q) — is in NP but widely believed not to be in P.
The resulting activation verbalization (best-of-N, reranked by AR fidelity) surfaced:
“This article originally appeared in the August 2016 edition of CACM. A new method of proving computational hardness of problems, known as multilinearization, can improve efficiency, reduce complexity and simplify proofs. In this article, I describe multilinearization and its application to several key problems, from the discrete logarithm and factoring to RSA and elliptic-curve discrete logarithms.”
What emerges is not a literal restatement, but a structured articulation of the model’s internal associations: hardness proofs, algebraic techniques, and the cryptographic implications that orbit this foundational question in computational complexity.
The demo offers a compelling interface for exploring these latent representations.
Explore it here:
RiverRider/srt-nla-av-v1-demo
Recommended: Best-of-N sampling with round-trip evaluation for highest fidelity.
I introduced a concise mathematical formulation of the P versus NP question into the SRT-NLA-AV-v1 demonstration:
P vs NP asks whether every problem whose solution can be verified in polynomial time (NP) can also be solved in polynomial time (P). Integer factorization — given N = p·q where p and q are large primes (p < q) — is in NP but widely believed not to be in P.
The resulting activation verbalization (best-of-N, reranked by AR fidelity) surfaced:
“This article originally appeared in the August 2016 edition of CACM. A new method of proving computational hardness of problems, known as multilinearization, can improve efficiency, reduce complexity and simplify proofs. In this article, I describe multilinearization and its application to several key problems, from the discrete logarithm and factoring to RSA and elliptic-curve discrete logarithms.”
What emerges is not a literal restatement, but a structured articulation of the model’s internal associations: hardness proofs, algebraic techniques, and the cryptographic implications that orbit this foundational question in computational complexity.
The demo offers a compelling interface for exploring these latent representations.
Explore it here:
RiverRider/srt-nla-av-v1-demo
Recommended: Best-of-N sampling with round-trip evaluation for highest fidelity.
- 1 reply
liked a
model 10 days ago
reacted to salma-remyx's post with 🔥 12 days ago
Post
11574
The space of possible improvements for your AI model is large while evaluation is costly.
So I was excited to discover the ICML 2026 paper from Kobalczyk, Lin, Letham, Zhao, Balandat, and Bakshy titled "LILO: Bayesian Optimization with Natural Language Feedback."
The method learns efficiently from expert preferences, balancing exploration and exploitation in a principled way with Bayesian Optimization for expensive-to-evaluate black-box objectives.
Experimenting with the technique, I trained a Gaussian Process proxy model on the implicit preferences in my code repo's commit history at VQASynth.
The result: I used the model's preference scores to re-rank candidate papers recommended based on my interests in spatial reasoning and multimodal data synthesis.
Semantic relevance is a high-recall method for finding arXiv papers personalized to your interests. Adding contributor preferences, extracted from the merge history of your code offers a high-precision filter.
So what's next? I'm using the model to synthesize a larger volume of preference data to finetune an open-weight coding model with DPO and LoRA. Tuning Coding Agents via Implicit Preference Distillation
arXiv: https://arxiv.org/pdf/2510.17671
Substack: https://remyxai.substack.com/p/lilo-and-myx
VQASynth: https://github.com/remyxai/VQASynth
So I was excited to discover the ICML 2026 paper from Kobalczyk, Lin, Letham, Zhao, Balandat, and Bakshy titled "LILO: Bayesian Optimization with Natural Language Feedback."
The method learns efficiently from expert preferences, balancing exploration and exploitation in a principled way with Bayesian Optimization for expensive-to-evaluate black-box objectives.
Experimenting with the technique, I trained a Gaussian Process proxy model on the implicit preferences in my code repo's commit history at VQASynth.
The result: I used the model's preference scores to re-rank candidate papers recommended based on my interests in spatial reasoning and multimodal data synthesis.
Semantic relevance is a high-recall method for finding arXiv papers personalized to your interests. Adding contributor preferences, extracted from the merge history of your code offers a high-precision filter.
So what's next? I'm using the model to synthesize a larger volume of preference data to finetune an open-weight coding model with DPO and LoRA. Tuning Coding Agents via Implicit Preference Distillation
arXiv: https://arxiv.org/pdf/2510.17671
Substack: https://remyxai.substack.com/p/lilo-and-myx
VQASynth: https://github.com/remyxai/VQASynth
- 1 reply
liked 4
models 13 days ago
liked 2
models 14 days ago
liked a
model 16 days ago