Downloads · 30 days
14
10% of all-time downloads
AmanPriyanshu/gpt-oss-14.3b-specialized-science-pruned-moe-only-21-experts
gpt-oss-14.3b-specialized-science-pruned-moe-only-21-experts is a text generation model from AmanPriyanshu. Use it when you need the model to write or continue text. The card lists the license as apache-2.0.
Project: https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/
Downloads · 30 days
14
10% of all-time downloads
All-time downloads
134
Public
Parameters
14.3B
28.7 GB on disk
Likes
1
Public
Click a slice to open those files.
.safetensors28.7 GB · 100%
From the Hugging Face model README
Project: https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/
<div align="center">This is a pruned variant of OpenAI's GPT-OSS-20B model, reduced to 21 experts per layer based on activation patterns from the AmanPriyanshu/GPT-OSS-20B MoE Expert Activations dataset. We analyzed router decisions across evaluation benchmarks to identify and retain experts most relevant for science tasks.
⚠️ Experimental Model: This is an experimental pruned model that may not work well - check the examples below to see if the outputs meet your needs before use.
This pruning approach reduces the model size while attempting to preserve performance on the target domain.
| Metric | Value |
|---|---|
| Base Model | openai/gpt-oss-20b |
| Architecture | Mixture-of-Experts Transformer |
| Total Parameters | ~14.3B (pruned from 21B) |
| Original Experts per Layer | 32 |
| Pruned Experts per Layer | 21 |
| Layers | 24 |
| Top-k Routing | 4 |
| Context Length | 128K tokens |
| Attention Heads | 64 (Query), 8 (Key-Value) |
| Residual Dimension | 2880 |
| Attention Pattern | Alternating dense & sliding window (128 tokens) |
| Positional Encoding | RoPE (Rotary Position Embedding) |
| Normalization | RMSNorm |
| Precision | BF16 |
| License | Apache 2.0 |
| Specialization | Science |
Mixture-of-Experts models contain multiple specialized sub-networks (experts) per layer. During inference, only a subset of experts are activated for each token. Expert pruning involves:
Note: Performance may vary depending on how well the pruned experts match your specific use case.
This science-specialized model leverages experts that showed high activation patterns during scientific reasoning tasks from GPQA (physics, chemistry, biology) and MMLU science domains. These experts demonstrate superior performance on complex scientific problem-solving and technical knowledge recall.
The expert selection process utilized our comprehensive analysis of router activation patterns across multiple evaluation benchmarks:
By identifying experts that consistently activated for science tasks, we created this specialized model that maintains domain expertise while significantly reducing computational requirements from 32 to 21 experts per layer.
This model is based on analysis from the GPT-OSS-20B MoE Expert Activations dataset available at: 🔗 https://huggingface.co/datasets/AmanPriyanshu/GPT-OSS-20B-MoE-expert-activations
The dataset contains router activation patterns from OpenAI's GPT-OSS-20B model across diverse evaluation benchmarks, enabling the creation of these domain-optimized models through systematic expert pruning.
Our approach involves:
This is a direct pruning approach - no additional training was performed. The model inherits all capabilities from the original GPT-OSS-20B with focused expert selection.
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
# Load the specialized model on CPU
model = AutoModelForCausalLM.from_pretrained(
"AmanPriyanshu/gpt-oss-14.3b-specialized-science-pruned-moe-only-21-experts",
torch_dtype=torch.bfloat16,
device_map="cpu",
trust_remote_code=True
)
tokenizer = AutoTokenizer.from_pretrained("AmanPriyanshu/gpt-oss-14.3b-specialized-science-pruned-moe-only-21-experts")
# Generate with the model
messages = [
{"role": "user", "content": "Explain the process of photosynthesis in plants."}
]
inputs = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
return_tensors="pt",
return_dict=True,
reasoning_effort="medium"
)
# Ensure inputs are on the same device as model
inputs = {k: v.to(model.device) for k, v in inputs.items()}
outputs = model.generate(
**inputs,
max_new_tokens=512,
do_sample=True,
temperature=0.1,
top_p=0.9,
pad_token_id=tokenizer.eos_token_id,
eos_token_id=tokenizer.eos_token_id
)
# Decode only the generated part
input_length = inputs['input_ids'].shape[1]
response_tokens = outputs[0][input_length:]
response = tokenizer.decode(response_tokens, skip_special_tokens=True)
print(response)
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
# Check MPS availability and load model
device = "mps" if torch.backends.mps.is_available() else "cpu"
model = AutoModelForCausalLM.from_pretrained(
"AmanPriyanshu/gpt-oss-14.3b-specialized-science-pruned-moe-only-21-experts",
torch_dtype=torch.float16, # Better MPS compatibility
device_map=device,
trust_remote_code=True,
low_cpu_mem_usage=True
)
tokenizer = AutoTokenizer.from_pretrained("AmanPriyanshu/gpt-oss-14.3b-specialized-science-pruned-moe-only-21-experts")
# Generate with the model
messages = [
{"role": "user", "content": "Explain the process of photosynthesis in plants."}
]
inputs = tokenizer.apply_chat_template(
messages,
add_generation_prompt=True,
return_tensors="pt",
return_dict=True,
reasoning_effort="medium"
)
# Move inputs to model device
inputs = {k: v.to(model.device) if isinstance(v, torch.Tensor) else v for k, v in inputs.items()}
# Use torch.no_grad for MPS stability
with torch.no_grad():
outputs = model.generate(
**inputs,
max_new_tokens=512,
do_sample=True,
temperature=0.1,
top_p=0.9,
pad_token_id=tokenizer.eos_token_id,
eos_token_id=tokenizer.eos_token_id,
use_cache=True
)
# Decode only the generated part
input_length = inputs['input_ids'].shape[1]
response_tokens = outputs[0][input_length:]
response = tokenizer.decode(response_tokens, skip_special_tokens=True)
print(response)
For CUDA GPU inference, the models work with minimal modifications from the CPU example above. Simply change:
device_map="auto" # Will automatically use GPU if available
torch_dtype=torch.bfloat16 # or torch.float16
The following examples demonstrate the model's performance on science tasks:
Question: Explain the process of photosynthesis in plants.
Response:
<|channel|>analysis<|message|>Need concise explanation.<|end|><|start|>assistant<|channel|>final<|message|>**Photosynthesis in Plants – A Step‑by‑Step Overview**
| **Stage** | **What Happens** | **Key Players** | **Result** |
|---|---|---|---|
| 1. **Light Capture** | Sunlight hits chlorophyll‑rich cells in the leaf’s chloroplasts. | • Chlorophyll a & b (pigments) | • Excited electrons are released from chlorophyll |
| 2. **Electron Transport Chain (ETC)** | Excited electrons travel through a series of proteins in the thylakoid membrane, pumping protons into the thylakoid lumen. | • Photosystem II (PSII) → Photosystem I (PSI) → cytochrome b₆f complex → ATP synthase | • Generates a proton gradient → ATP (via chemiosmosis) |
| 3. **Water Splitting (Oxygen Evolution)** | PSII uses light energy to split H₂O into O₂, 4H⁺, and 4e⁻. | • Oxygenase enzyme (part of PSII) | • O₂ released to atmosphere; electrons feed ETC |
| 4. **NADPH Formation** | Electrons reduce NADP⁺ to NADPH at PSI. | • PSI, ferredoxin, NADP⁺ reductase | • NADPH carries reducing power |
| 5. **Carbon Fixation (Calvin Cycle)** | CO₂ from the air is fixed into sugars using ATP and NADPH. | • Ribulose‑1‑5‑bisphosphate carboxylase/oxygenase (RuBisCO) | • Produces 3‑phosphoglycerate → glyceraldehyde‑3‑phosphate → glucose & other carbohydrates |
| 6. **Regeneration of CO₂ Carrier** | 3‑phosphoglycerate is recycled back to ribulose‑1‑5‑bisphosphate, completing the cycle. | • Various enzymes (e.g., GAPDH, PGK, aldolase) | • Cycle repeats, consuming more CO₂ and light |
### Quick Summary
1. **Light energy** excites electrons in chlorophyll.
2. Electrons travel through the ETC, creating a proton gradient that powers **ATP synthesis**.
3. **Water is split**, releasing O₂ and providing electrons.
4. Electrons reduce NADP⁺ to
Question: What causes the greenhouse effect and how does it work?
Response:
<|channel|>analysis<|message|>Need explanation.<|end|><|start|>assistant<|channel|>final<|message|>### What Causes the Greenhouse Effect?
The greenhouse effect is a natural process that keeps Earth’s surface warm enough to support life. It is caused by certain gases in the atmosphere—called **greenhouse gases (GHGs)**—that absorb, trap, and re‑emit infrared (IR) radiation emitted by the planet. The main greenhouse gases are:
| Gas | Typical atmospheric concentration | Primary source |
|-----|-----------------------------------|-----------------|
| **Water vapor (H₂O)** | ~0.3 % (variable) | Evaporation from oceans, soil, plants |
| **Carbon dioxide (CO₂)** | ~0.04 % (≈ 400 ppm) | Fossil‑fuel combustion, deforestation, volcanic activity |
| **Methane (CH₄)** | ~0.00018 % (≈ 1.8 ppm) | Agriculture (livestock, rice paddies), wetlands, fossil‑fuel extraction |
| **Nitrous oxide (N₂O)** | ~0.00003 % (≈ 0.3 ppm) | Agriculture (fertilizers), industrial processes |
| **Ozone (O₃)** | ~0.00002 % (≈ 0.2 ppm) | Photochemical reactions in the stratosphere |
| **CFCs, HCFCs, HFCs, PFCs** | trace | Industrial refrigerants, foam blowing agents, etc. |
These gases are **transparent to visible sunlight** (the short‑wave radiation that reaches Earth from the Sun) but **opaque to the long‑wave (infrared) radiation** that Earth emits after absorbing that sunlight. That opacity is what creates the greenhouse effect.
---
### How the Greenhouse Effect Works (Step‑by‑Step)
1. **Solar Radiation Reaches Earth**
- The Sun emits a spectrum of electromagnetic radiation.
- About **30 %** of the incoming solar energy is reflected back to space by clouds, ice, and the atmosphere.
- The remaining **70 %** is absorbed by the Earth’s surface (land, oceans, vegetation).
2. **Earth’s Surface Heats Up**
- The absorbed solar energy warms the surface, which then emits thermal radiation (infrared, IR) at wavelengths around 5–15 µm.
3. **Greenhouse Gases Absorb IR**
Question: Describe the structure and function of DNA.
Response:
<|channel|>analysis<|message|>Need concise description.<|end|><|start|>assistant<|channel|>final<|message|>**DNA (Deoxyribonucleic Acid)** is the hereditary material that carries the genetic instructions for all living organisms and many viruses. Its structure and function can be understood in two complementary ways: the *molecular architecture* that makes it a stable, information‑carrying polymer, and the *biological roles* it plays in cells.
---
## 1. Structural Features
| Feature | Description | Biological Significance |
|---|---|---|
| **Double‑helix backbone** | Two strands of nucleotides wound around each other in a right‑handed helix. Each strand is a sugar‑phosphate chain (deoxyribose + phosphate). | Provides mechanical stability and protects the genetic code from damage. |
| **Base pairing (Watson–Crick)** | Adenine (A) pairs with Thymine (T) via two hydrogen bonds; Cytosine (C) pairs with Guanine (G) via three hydrogen bonds. | Enables accurate copying of genetic information during replication. |
| **Antiparallel orientation** | One strand runs 5′→3′, the other 3′→5′. | Required for enzymes (DNA polymerases, helicases) to read and synthesize strands correctly. |
| **Major & minor grooves** | The helical twist creates wide (major) and narrow (minor) grooves. | Sites where proteins (transcription factors, nucleases) bind to regulate gene expression. |
| **Base composition & GC content** | GC pairs are stronger (3 H‑bonds) than AT pairs. | Influences DNA stability, melting temperature, and chromatin structure. |
| **Chromatin packaging** | In eukaryotes, DNA wraps around histone octamers to form nucleosomes, then higher‑order fibers. | Condenses DNA into chromosomes while allowing regulated access to genes. |
| **Topological properties** | DNA can be super‑coiled, knotted, or branched. | Facilitates compaction and influences transcription/replication dynamics. |
---
## 2. Functional Roles
| Function | How DNA accomplishes it | Key Players |
|---|---|---|
| **Genetic information storage** | Sequence of bases encodes genes (protein‑coding, regulatory, non‑coding). | Nucleotides, base‑pairing rules |
| **Replication** | DNA polymerases synthesize
If you use this model in your research, please cite:
@misc{priyanshu2025gptoss,
title={{GPT-OSS MoE Expert Fingerprinting: Analyzing Expert Activation Patterns in Mixture of Experts Models}},
author={Priyanshu, Aman and Vijay, Supriti},
year={2025},
howpublished={\url{https://amanpriyanshu.github.io/GPT-OSS-MoE-ExpertFingerprinting/}},
note={Interactive analysis tool for expert activation patterns in MoE architectures}
}