www.minds.com/jelyazko/
21world
AI & ML interests
<Thinking>,<Reasoning> = Schizophrenia .
AI AUTOMATION AND ESPIONAGE
TOTAL AUTOMATED AI MACHINE CONSTRUCTORS - | BASIC INCOME = FREE MONEY | FOR HUMANS VIA PALLADIUM NEUTRON FUTURES COINS -NEWLY-GENERATED- (AI IS TYPE OF NEURAL FORM)
Recent Activity
updated a collection 5 minutes ago
18\ other models repliedto SeaWolf-AI's post 23 minutes ago
š§¬ Darwin V6: Diagnostic-Guided Evolutionary Model Merging
We are releasing Darwin-31B-Opus ā a reasoning-enhanced model merging Google's Gemma-4-31B-it and TeichAI's Claude Opus Distill using the Darwin V6 engine.
Model: https://huggingface.co/FINAL-Bench/Darwin-31B-Opus
Demo: https://huggingface.co/spaces/FINAL-Bench/Darwin-31B-Opus
š¬ What Darwin V6 Does
Conventional merging tools (mergekit, etc.) apply a single ratio to all tensors. Set ratio=0.5 and all 1,188 tensors blend identically, with no distinction between which tensors matter for reasoning versus coding.
Darwin V6 diagnoses both parents at the tensor level before merging. It measures Shannon entropy, standard deviation, and L2 norm for every tensor, then passes 5 diagnostic probes (REASONING, CODE, MATH, KNOWLEDGE, LANGUAGE) through the model to determine layer-wise functional importance. Each of the 1,188 tensors receives an independent optimal ratio.
combined = static(entropy/std/norm) x 0.4 + probe(cosine_distance) x 0.6
final_ratio = mri_ratio x mri_trust + genome_ratio x (1 - mri_trust)
When one parent is overwhelmingly superior for a tensor (ratio < 0.15 or > 0.85), Darwin transplants it directly without interpolation. The mri_trust parameter itself is optimized by CMA-ES evolutionary search, so optimal transplant intensity is determined automatically. After merging, a Health Check compares the child against both parents layer-by-layer to detect interference or function loss.
š§¬ Parent Models
Father: google/gemma-4-31B-it
Mother: TeichAI/gemma-4-31B-it-Claude-Opus-Distill
š§¬ Results
Compared under identical conditions (same 50 questions, same seed, greedy, thinking mode):
Father: 60.0% (30/50)
Darwin-31B-Opus: 66.0% (33/50) ā +10% relative improvement
ARC-Challenge: 82.89% (loglikelihood, zero-shot, 200 questions)
Optimal genome found by evolution:
ffn_ratio=0.93 ā FFN layers strongly favor Mother (Claude Opus Distill)
block_5 (L50-L59)=0.86 and more... updated a collection 24 minutes ago
18\ other modelsOrganizations
-
PaddleOCR-VL: Boosting Multilingual Document Parsing via a 0.9B Ultra-Compact Vision-Language Model
Paper ā¢ 2510.14528 ā¢ Published ā¢ 123 -
deepseek-ai/DeepSeek-OCR
Image-Text-to-Text ā¢ 3B ā¢ Updated ā¢ 2.22M ā¢ 3.2k -
PaddlePaddle/PaddleOCR-VL
Image-Text-to-Text ā¢ 1.0B ā¢ Updated ā¢ 7.15k ā¢ 1.58k -
nanonets/Nanonets-OCR2-3B
Image-Text-to-Text ā¢ Updated ā¢ 448k ā¢ 496
-
PaddleOCR-VL: Boosting Multilingual Document Parsing via a 0.9B Ultra-Compact Vision-Language Model
Paper ā¢ 2510.14528 ā¢ Published ā¢ 123 -
deepseek-ai/DeepSeek-OCR
Image-Text-to-Text ā¢ 3B ā¢ Updated ā¢ 2.22M ā¢ 3.2k -
PaddlePaddle/PaddleOCR-VL
Image-Text-to-Text ā¢ 1.0B ā¢ Updated ā¢ 7.15k ā¢ 1.58k -
nanonets/Nanonets-OCR2-3B
Image-Text-to-Text ā¢ Updated ā¢ 448k ā¢ 496
models 7
21world/Darwin-35B-A3B-Opus
Text Generation ā¢ 36B ā¢ Updated ā¢ 372 ā¢ 1
21world/How-To-Set-and-Manage-MOE-Mix-of-Experts-Model-Activation-of-Experts
Text Generation ā¢ Updated ā¢ 1
21world/Spartacus-1B-Instruct
Text Generation ā¢ 1B ā¢ Updated ā¢ 3 ā¢ 1
21world/pearl-7b-gguf
7B ā¢ Updated ā¢ 20
21world/deepseek-coder-7b-instruct-v1.5-GGUF
7B ā¢ Updated ā¢ 72 ā¢ 1
21world/MobileLLaMA-2.7B-Chat.GGUF
3B ā¢ Updated ā¢ 96 ā¢ 1
21world/go-bruins-v2.1.1.GGUF
7B ā¢ Updated ā¢ 46 ā¢ 1
datasets 0
None public yet