Article
Training-Free Reasoning at 88.89% on GPQA Diamond: How Darwin Family Hit Frontier Scores Without a Single Gradient Step
FINAL-Bench
• • 11Minseo Kim
MinseoKim-03
AI & ML interests
None yet
Recent Activity
reacted to SeaWolf-AI's post with 🔥 about 10 hours ago
🧬 Darwin Family: Zero Gradient Steps, GPQA Diamond 88.89%
How far can we push LLM reasoning *without* training?
Our team at VIDRAFT submitted this paper to Daily Papers yesterday, and it's
currently #3. Huge thanks to everyone who upvoted — sharing the core ideas below.
🔗 Paper: https://huggingface.co/papers/2605.14386
🔗 arXiv: https://arxiv.org/abs/2605.14386
🔗 Model: https://huggingface.co/FINAL-Bench/Darwin-28B-Opus
---
TL;DR
Darwin Family is a training-free evolutionary merging framework.
By recombining the weight spaces of existing LLM checkpoints — with zero
gradient-based training — it reaches frontier-level reasoning.
- 🏆 Darwin-28B-Opus: GPQA Diamond 88.89%
- 💸 Zero gradient steps — not a single B200 or H200 hour needed
- 🧬 Consistent gains across 4B → 35B scale
- 🔀 Cross-architecture breeding between Transformer and Mamba families
- 🔁 Stable recursive multi-generation evolution
#Three Core Mechanisms
① 14-dim Adaptive Merge Genome — fine-grained recombination at both
component level (Attention / FFN / MLP / LayerNorm / Embedding) and block
level, expanding the prior evolutionary-merge search space.
② MRI-Trust Fusion — we diagnose each layer's reasoning contribution
via an **MRI (Model Reasoning Importance)** signal and fuse it with
evolutionary search through a **learnable trust parameter**. Trust the
diagnostic too much and search collapses; ignore it and search becomes
inefficient — Darwin learns the balance from data.
③ Architecture Mapper — weight-space breeding across heterogeneous
families. Attention × SSM crossover actually works.
Why It Matters
> Diagnose latent capabilities already encoded in open checkpoints,
> and recombine them — no gradients required.
Replies and critiques welcome 🙌 upvoted a paper about 11 hours ago
Darwin Family: MRI-Trust-Weighted Evolutionary Merging for Training-Free Scaling of Language-Model ReasoningOrganizations
None yet
upvoted an article about 10 hours ago
reacted to SeaWolf-AI's post with 🔥 about 10 hours ago
Post
898
🧬 Darwin Family: Zero Gradient Steps, GPQA Diamond 88.89%
How far can we push LLM reasoning *without* training?
Our team at VIDRAFT submitted this paper to Daily Papers yesterday, and it's
currently #3. Huge thanks to everyone who upvoted — sharing the core ideas below.
🔗 Paper: Darwin Family: MRI-Trust-Weighted Evolutionary Merging for Training-Free Scaling of Language-Model Reasoning (2605.14386)
🔗 arXiv: https://arxiv.org/abs/2605.14386
🔗 Model: FINAL-Bench/Darwin-28B-Opus
---
TL;DR
Darwin Family is a training-free evolutionary merging framework.
By recombining the weight spaces of existing LLM checkpoints — with zero
gradient-based training — it reaches frontier-level reasoning.
- 🏆 Darwin-28B-Opus: GPQA Diamond 88.89%
- 💸 Zero gradient steps — not a single B200 or H200 hour needed
- 🧬 Consistent gains across 4B → 35B scale
- 🔀 Cross-architecture breeding between Transformer and Mamba families
- 🔁 Stable recursive multi-generation evolution
#Three Core Mechanisms
① 14-dim Adaptive Merge Genome — fine-grained recombination at both
component level (Attention / FFN / MLP / LayerNorm / Embedding) and block
level, expanding the prior evolutionary-merge search space.
② MRI-Trust Fusion — we diagnose each layer's reasoning contribution
via an **MRI (Model Reasoning Importance)** signal and fuse it with
evolutionary search through a **learnable trust parameter**. Trust the
diagnostic too much and search collapses; ignore it and search becomes
inefficient — Darwin learns the balance from data.
③ Architecture Mapper — weight-space breeding across heterogeneous
families. Attention × SSM crossover actually works.
Why It Matters
> Diagnose latent capabilities already encoded in open checkpoints,
> and recombine them — no gradients required.
Replies and critiques welcome 🙌
How far can we push LLM reasoning *without* training?
Our team at VIDRAFT submitted this paper to Daily Papers yesterday, and it's
currently #3. Huge thanks to everyone who upvoted — sharing the core ideas below.
🔗 Paper: Darwin Family: MRI-Trust-Weighted Evolutionary Merging for Training-Free Scaling of Language-Model Reasoning (2605.14386)
🔗 arXiv: https://arxiv.org/abs/2605.14386
🔗 Model: FINAL-Bench/Darwin-28B-Opus
---
TL;DR
Darwin Family is a training-free evolutionary merging framework.
By recombining the weight spaces of existing LLM checkpoints — with zero
gradient-based training — it reaches frontier-level reasoning.
- 🏆 Darwin-28B-Opus: GPQA Diamond 88.89%
- 💸 Zero gradient steps — not a single B200 or H200 hour needed
- 🧬 Consistent gains across 4B → 35B scale
- 🔀 Cross-architecture breeding between Transformer and Mamba families
- 🔁 Stable recursive multi-generation evolution
#Three Core Mechanisms
① 14-dim Adaptive Merge Genome — fine-grained recombination at both
component level (Attention / FFN / MLP / LayerNorm / Embedding) and block
level, expanding the prior evolutionary-merge search space.
② MRI-Trust Fusion — we diagnose each layer's reasoning contribution
via an **MRI (Model Reasoning Importance)** signal and fuse it with
evolutionary search through a **learnable trust parameter**. Trust the
diagnostic too much and search collapses; ignore it and search becomes
inefficient — Darwin learns the balance from data.
③ Architecture Mapper — weight-space breeding across heterogeneous
families. Attention × SSM crossover actually works.
Why It Matters
> Diagnose latent capabilities already encoded in open checkpoints,
> and recombine them — no gradients required.
Replies and critiques welcome 🙌
upvoted a paper about 11 hours ago
liked a
model 20 days ago
liked a
model 21 days ago
reacted to SeaWolf-AI's post with 🔥 21 days ago
Post
8731
🧬 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
liked a
model 21 days ago