OzTianlu (Zixi "Oz" Li)

reacted to their post with 🔥 4 days ago

Post

2755

Geilim-1B-SR-Instruct — Serbian Intelligence for Deep Reasoning 🧠🇷🇸
NoesisLab/Geilim-1B-SR-Instruct
Geilim-1B-SR-Instruct is a lightweight Large Language Model (LLM) designed to bring advanced reasoning capabilities to low-resource languages. It focuses on Serbian understanding and generation while maintaining robust English reasoning. Built on the LLaMA-3 architecture with a proprietary hybrid reasoning mechanism, it delivers deep logic while keeping outputs concise and natural. 🚀

Core Innovations 💡

Implicit Deep Reasoning: Combines standard attention mechanisms with graph-structured reasoning components for rigorous logic and causal inference. 🕸️

ASPP & -flow Hybrid Design: High-efficiency structured propagation + internal probability space optimization for high-quality reasoning without long-winded intermediate steps. ⚡
Bilingual Adaptation: Primarily focused on Serbian while preserving English logic, making it perfect for multilingual chats and cross-lingual tasks. 🌍
Lightweight & Efficient: At ~1.3B parameters, it runs smoothly on consumer-grade GPUs, ideal for edge devices and research. 💻

Use Cases 🛠️

Serbian Chatbots: Intelligent assistants with local linguistic nuance. 🗣️
Educational Tools: Multi-turn interactive tasks and learning support. 📚

Key Advantages ✨

Clean Output: Avoids messy "thinking" tags; reasoning happens internally, delivering clear and direct results. ✅
Open Access: Licensed under Apache-2.0, making it easy for research and engineering integration. 🔓
AI Democratization: Empowering low-resource language ecosystems with cutting-edge intelligence. 🤝

1 reply

·

posted an update 4 days ago

Post

2755

Geilim-1B-SR-Instruct — Serbian Intelligence for Deep Reasoning 🧠🇷🇸
NoesisLab/Geilim-1B-SR-Instruct
Geilim-1B-SR-Instruct is a lightweight Large Language Model (LLM) designed to bring advanced reasoning capabilities to low-resource languages. It focuses on Serbian understanding and generation while maintaining robust English reasoning. Built on the LLaMA-3 architecture with a proprietary hybrid reasoning mechanism, it delivers deep logic while keeping outputs concise and natural. 🚀

Core Innovations 💡

Implicit Deep Reasoning: Combines standard attention mechanisms with graph-structured reasoning components for rigorous logic and causal inference. 🕸️

ASPP & -flow Hybrid Design: High-efficiency structured propagation + internal probability space optimization for high-quality reasoning without long-winded intermediate steps. ⚡
Bilingual Adaptation: Primarily focused on Serbian while preserving English logic, making it perfect for multilingual chats and cross-lingual tasks. 🌍
Lightweight & Efficient: At ~1.3B parameters, it runs smoothly on consumer-grade GPUs, ideal for edge devices and research. 💻

Use Cases 🛠️

Serbian Chatbots: Intelligent assistants with local linguistic nuance. 🗣️
Educational Tools: Multi-turn interactive tasks and learning support. 📚

Key Advantages ✨

Clean Output: Avoids messy "thinking" tags; reasoning happens internally, delivering clear and direct results. ✅
Open Access: Licensed under Apache-2.0, making it easy for research and engineering integration. 🔓
AI Democratization: Empowering low-resource language ecosystems with cutting-edge intelligence. 🤝

1 reply

·

reacted to yuriyvnv's post with 👍 4 days ago

Post

2142

🎯 WAVe: 1B Multimodal Embedding Model for Word-Level Speech Quality

Multimodal embeddings for speech + transcript that verify quality at the word level, not just sentence level. Catches mispronunciations, timing errors, and prosody issues that sentence-level filters miss.

📊 Impact on Portuguese ASR:
• 34% reduction in training steps
• 50% better cross-domain generalization
• 30% less synthetic data needed
• Word-aligned attention finds errors other methods miss

🏗️ Architecture:
• Text: XLM-RoBERTa (278M params)
• Audio: Wav2Vec2-BERT 2.0 (581M params)
• Word Alignment: Multi-head attention + GLU (14M params)
• Total: 1B parameters

from transformers import AutoModel, AutoProcessor

  processor = AutoProcessor.from_pretrained(
      "yuriyvnv/WAVe-1B-Multimodal-PT",
      trust_remote_code=True
  )
  model = AutoModel.from_pretrained(
      "yuriyvnv/WAVe-1B-Multimodal-PT",
      trust_remote_code=True
  )

# Assess speech-transcript alignment

inputs = processor(text="Olá, como está?", audio=audio_array, sampling_rate=16000, return_tensors="pt")
  quality = model(**inputs).quality_score.item()

Perfect for filtering synthetic speech datasets before ASR training.

Model: yuriyvnv/WAVe-1B-Multimodal-PT
Code to create WAVe : https://github.com/yuriyvnv/WAVe
#multimodal #speech #embeddings #asr
#syntheticdata #qualityassessment

1 reply

·

replied to their post 6 days ago

Great observation! You nailed it with the comment. 💗To clarify, when I mentioned this as an 'alternative,' I was referring to the implementation method (using a seamless pipeline without enforced tags), not necessarily a breakthrough in arithmetic capability at the 1B scale. What you're seeing here is a classic example of hallucination in small-parameter models. The model is faithfully following the instruction to 'reason step-by-step' (CoT), but due to its limited size (1B), it hallucinates the intermediate calculations while maintaining a confident tone. Maintaining logic while ensuring factual accuracy in such compact models is indeed one of the biggest challenges we are currently facing and working to optimize.

reacted to their post with 🚀 7 days ago

Post

2547

🚀 Geilim-1B-Instruct — Implicit Deep Reasoning, Zero Verbosity
NoesisLab/Geilim-1B-Instruct
https://huggingface.co/collections/NoesisLab/geilim-large-language-models
No <think> tags. No long CoT.
Reasoning happens inside the hidden states, not in the output.
What’s different
🧠 Implicit reasoning: deep causal reasoning without exposing chains
🕸️ ASPP (Adjacency-Structured Parallel Propagation): parent-only causal graph, O(n) message passing
🌊 π-flow: internal probability-space refinement instead of token-level deliberation
⚖️ Hybrid gating: learns when to use structure vs attention
Why it matters
Lower latency & token cost
Cleaner, production-ready outputs
CoT-level reasoning depth without verbosity tax
Built on Llama-3.2-1B-Instruct, trained for math, logic, and commonsense.
Designed for small-model reasoning at the edge.
#ImplicitReasoning #SmallLLM #EfficientAI #ReasoningModels #ASPP #PiFlow

2 replies

·

posted an update 7 days ago

Post

2547

🚀 Geilim-1B-Instruct — Implicit Deep Reasoning, Zero Verbosity
NoesisLab/Geilim-1B-Instruct
https://huggingface.co/collections/NoesisLab/geilim-large-language-models
No <think> tags. No long CoT.
Reasoning happens inside the hidden states, not in the output.
What’s different
🧠 Implicit reasoning: deep causal reasoning without exposing chains
🕸️ ASPP (Adjacency-Structured Parallel Propagation): parent-only causal graph, O(n) message passing
🌊 π-flow: internal probability-space refinement instead of token-level deliberation
⚖️ Hybrid gating: learns when to use structure vs attention
Why it matters
Lower latency & token cost
Cleaner, production-ready outputs
CoT-level reasoning depth without verbosity tax
Built on Llama-3.2-1B-Instruct, trained for math, logic, and commonsense.
Designed for small-model reasoning at the edge.
#ImplicitReasoning #SmallLLM #EfficientAI #ReasoningModels #ASPP #PiFlow

2 replies

·

replied to sergiopaniego's post 8 days ago

awesome!

reacted to sergiopaniego's post with 🔥 8 days ago

Post

2483

New TRL + OpenEnv example! 💥

Fine tune an LLM for playing Sudoku using an RL env via OpenEnv

Includes a script that runs on 1 or multiple GPUs with vLLM, plus a Colab-ready notebook.

Enjoy!

Notebook: https://colab.research.google.com/github/huggingface/trl/blob/main/examples/notebooks/openenv_sudoku_grpo.ipynb

Script: https://github.com/huggingface/trl/blob/main/examples/scripts/openenv/sudoku.py

1 reply

·

reacted to imnotkitty's post with 🔥 8 days ago

Post

2298

📌Same day, Two Releases.

Jan 27th just got interesting on Open-source AI modles.
✅Kimi K2.5: How to make models "think" across text and vision natively?
moonshotai/Kimi-K2.5
✅DeepSeek-OCR 2: How to make models "see" more like humans, not scanners?
deepseek-ai/DeepSeek-OCR-2

One focuses on depth of reasoning, the other on precision of vision.
What's the key differentiator for a multimodal model in your view: raw power or computational elegance?

reacted to their post with ❤️ 16 days ago

Post

1189

🚀 Introducing Asterisk — Hybrid ASPP-Attention Architecture! 🌟

https://huggingface.co/NoesisLab/Asterisk

We’re excited to launch Asterisk, a cutting-edge language model by NoesisLab on Hugging Face! 🎉 Built on top of SmolLM2-135M-Instruct, Asterisk integrates Adjacency-Structured Parallel Propagation (ASPP) with standard attention to bring structured reasoning power into language modeling.

✨ Key Highlights:

🔹 Hybrid Architecture – Fuses graph-centric ASPP local reasoning with global attention for richer representations.
🔹 Enhanced Reasoning – ASPP enables iterative local state evolution that complements traditional transformer layers.
🔹 Efficient Design – ~171M parameters with smart supervised fine-tuning (Capybara dataset).
🔹 Flexible & Open – Apache-2.0 licensed and ready to integrate via Hugging Face 🤗 Transformers.

📈 Asterisk showcases how hybrid operators — inspired by theoretical frameworks like the Asterisk Operator — can bring structured reasoning into modern LMs in a scalable way.

👉 Try it out, explore the code, and start building: huggingface.co/NoesisLab/Asterisk

1 reply

·

posted an update 17 days ago

Post

1189

🚀 Introducing Asterisk — Hybrid ASPP-Attention Architecture! 🌟

https://huggingface.co/NoesisLab/Asterisk

We’re excited to launch Asterisk, a cutting-edge language model by NoesisLab on Hugging Face! 🎉 Built on top of SmolLM2-135M-Instruct, Asterisk integrates Adjacency-Structured Parallel Propagation (ASPP) with standard attention to bring structured reasoning power into language modeling.

✨ Key Highlights:

🔹 Hybrid Architecture – Fuses graph-centric ASPP local reasoning with global attention for richer representations.
🔹 Enhanced Reasoning – ASPP enables iterative local state evolution that complements traditional transformer layers.
🔹 Efficient Design – ~171M parameters with smart supervised fine-tuning (Capybara dataset).
🔹 Flexible & Open – Apache-2.0 licensed and ready to integrate via Hugging Face 🤗 Transformers.

📈 Asterisk showcases how hybrid operators — inspired by theoretical frameworks like the Asterisk Operator — can bring structured reasoning into modern LMs in a scalable way.

👉 Try it out, explore the code, and start building: huggingface.co/NoesisLab/Asterisk

1 reply

·

Zixi "Oz" Li

AI & ML interests

Recent Activity

Organizations

Zixi "Oz" Li

AI & ML interests

Recent Activity

Organizations

OzTianlu's activity