FINAL-Bench/Darwin-28B-KR
Text Generation • 27B • Updated • 203 • 7
gen
ginigini
AI & ML interests
None yet
Recent Activity
liked a model 1 day ago
FINAL-Bench/Darwin-28B-KR liked a model 1 day ago
mradermacher/Darwin-9B-NEG-GGUF liked a model 1 day ago
mradermacher/Darwin-4B-Genesis-GGUFOrganizations
None yet
liked 16
models 1 day ago
mradermacher/Darwin-9B-NEG-GGUF
9B • Updated • 910 • 6
mradermacher/Darwin-4B-Genesis-GGUF
8B • Updated • 1.39k • 7
Cheuv/Darwin-4B-Genesis-Q8_0-GGUF
Text Generation • 8B • Updated • 73 • 8
AMAImedia/Qwen3.5-35B-A3B-Darwin-Opus-NOESIS-AWQ-INT8
Text Generation • 37B • Updated • 84 • 9
AMAImedia/Qwen3.5-35B-A3B-Darwin-Opus-NOESIS-AWQ-INT4
Text Generation • 20B • Updated • 100 • 9
AMAImedia/Qwen3-1.7B-TTS-Cross-Darwin-NOESIS-AWQ-INT4
Text-to-Speech • 2B • Updated • 32 • 12
AMAImedia/Qwen3.5-9B-Darwin-Opus-NOESIS-AWQ-INT4
Image-Text-to-Text • 9B • Updated • 534 • 9
TheCluster/Darwin-35B-A3B-Opus-MLX-mixed-9bit
Image-Text-to-Text • 35B • Updated • 120 • 9
TheCluster/Darwin-35B-A3B-Opus-MLX-mxfp8
Image-Text-to-Text • 35B • Updated • 144 • 10
TheCluster/Darwin-35B-A3B-Opus-MLX-bf16
Image-Text-to-Text • 35B • Updated • 206 • 11
Odiora/Darwin-4B-Genesis-Q4_K_M-GGUF
Text Generation • 8B • Updated • 415 • 11
mradermacher/Darwin-4B-David-i1-GGUF
8B • Updated • 3.31k • 10
AI-Joe-git/Darwin-9B-Opus-GGUF
Text Generation • 9B • Updated • 377 • 11
FINAL-Bench/Darwin-2B-Opus
Text Generation • 2B • Updated • 1.07k • 14
bartowski/FINAL-Bench_Darwin-36B-Opus-GGUF
Text Generation • 35B • Updated • 4.19k • 22
liked a
model 2 days ago
liked a
Space 3 days ago
liked a
model 3 days ago
reacted to SeaWolf-AI's post with 🔥 3 days ago
Post
8458
🧬 Introducing Darwin-9B-NEG — the first model with Native Entropy Gating (NEG)
🔗 Try it now: FINAL-Bench/Darwin-9B-NEG
🔗 Q4 bit : FINAL-Bench/Darwin-9B-MFP4
We're thrilled to release Darwin-9B-NEG, a 9B-parameter reasoning model
that embeds an architecturally-internalised sense of self-confidence directly
into the transformer — our proprietary Native Entropy Gating (NEG) technology.
📊 GPQA Diamond (198 PhD-level questions):
▸ Baseline Darwin-9B (no NEG) → 51.01 %
▸ Pure NEG (greedy · 1× cost) → 63.64 % 🔥 +12.63 %p
▸ + Permutation (4× cost) → 76.26 %
▸ + Ensemble Refinement (~20×) → 84.34 % 🏆
With only 9 billion parameters and 1× inference cost, Pure NEG jumps
+12.63 %p over the same model without NEG. Going all-in with ensemble
refinement pushes it to 84.34 % — surpassing the published Qwen3.5-9B
leaderboard score (81.7 %) by +2.64 %p.
🔬 What makes NEG different from Multi-Turn Iteration (MTI)?
Classical MTI needs 3-8× extra inference passes. NEG instead lives
INSIDE the single decoding loop. Two tiny modules ride with the
transformer: NEG-Head predicts per-token entropy from the last hidden
state, and NEG-Gate conditionally restricts the top-k choice when
confidence is low. The gate activates in only 4.36 % of tokens —
essentially free at inference time.
✨ Key differentiators
• Architecturally internalised — model file *is* the feature
• 1× inference cost (vs. 3-8× for MTI)
• Drop-in with vLLM / SGLang / TGI / transformers — no extra engine
• +12.63 %p reasoning at zero latency overhead
• Single-file deployment, Apache 2.0 licensed
🧬 Lineage
Qwen/Qwen3.5-9B → Darwin-9B-Opus (V7 evolutionary merge) → Darwin-9B-NEG (V8 + NEG training)
#Darwin #NEG #NativeEntropyGating #GPQA #Reasoning #LLM #OpenSource #Apache2
🔗 Try it now: FINAL-Bench/Darwin-9B-NEG
🔗 Q4 bit : FINAL-Bench/Darwin-9B-MFP4
We're thrilled to release Darwin-9B-NEG, a 9B-parameter reasoning model
that embeds an architecturally-internalised sense of self-confidence directly
into the transformer — our proprietary Native Entropy Gating (NEG) technology.
📊 GPQA Diamond (198 PhD-level questions):
▸ Baseline Darwin-9B (no NEG) → 51.01 %
▸ Pure NEG (greedy · 1× cost) → 63.64 % 🔥 +12.63 %p
▸ + Permutation (4× cost) → 76.26 %
▸ + Ensemble Refinement (~20×) → 84.34 % 🏆
With only 9 billion parameters and 1× inference cost, Pure NEG jumps
+12.63 %p over the same model without NEG. Going all-in with ensemble
refinement pushes it to 84.34 % — surpassing the published Qwen3.5-9B
leaderboard score (81.7 %) by +2.64 %p.
🔬 What makes NEG different from Multi-Turn Iteration (MTI)?
Classical MTI needs 3-8× extra inference passes. NEG instead lives
INSIDE the single decoding loop. Two tiny modules ride with the
transformer: NEG-Head predicts per-token entropy from the last hidden
state, and NEG-Gate conditionally restricts the top-k choice when
confidence is low. The gate activates in only 4.36 % of tokens —
essentially free at inference time.
✨ Key differentiators
• Architecturally internalised — model file *is* the feature
• 1× inference cost (vs. 3-8× for MTI)
• Drop-in with vLLM / SGLang / TGI / transformers — no extra engine
• +12.63 %p reasoning at zero latency overhead
• Single-file deployment, Apache 2.0 licensed
🧬 Lineage
Qwen/Qwen3.5-9B → Darwin-9B-Opus (V7 evolutionary merge) → Darwin-9B-NEG (V8 + NEG training)
#Darwin #NEG #NativeEntropyGating #GPQA #Reasoning #LLM #OpenSource #Apache2