AX Diagnostic Leaderboard
π©Ί
LLM AX diagnostic leaderboard across 7 axes
π In a Training Loop
VIDRAFT_LAB
SeaWolf-AI
AI & ML interests
Contact: arxivgpt@gmail.com
Recent Activity
updated a Space about 21 hours ago
FINAL-Bench/ax-diagnostic published a Space about 21 hours ago
FINAL-Bench/ax-diagnostic repliedto their post 3 days ago
π Adding a GPU without building one
AI is usually framed as "how smart is the model / how many GPUs did you buy." The real bottleneck is elsewhere β how efficiently you use the GPUs you already have.
Training happens once; inference runs the entire time users use your product. So a service's economics come down to cost per token. Inference acceleration uses software to pull several times more out of the same GPU β the effect of plugging in one more "virtual GPU."
VIDRAFT's VKAE, measured (B200, same-harness, no quality loss):
Qwen3.5-35B-A3B (MoE): 25.7 β 601 tok/s (23.4Γ)
Darwin-36B-Opus (in-house MoE): 25.0 β 280.8 (11.2Γ)
10,000+ tok/s peak aggregate under concurrency
The key: it's reproducible β model + serving shipped as one container.
docker pull vidraft/qwen35-vkae:601
Don't take our word for it β run it yourself. The mechanism will be released as a paper.
π Leaderboard & demo π https://huggingface.co/spaces/VIDraft/vkae
Articles π https://huggingface.co/blog/FINAL-Bench/vkae-leaderboardOrganizations
updated a
Space about 21 hours ago
published a
Space about 21 hours ago
replied to their post 3 days ago
Agreed on the framing, and the push is fair.
25.7 (single-stream) and 601 (aggregate under concurrency) aren't the same axis β you're right. The 23.4Γ is a system-level number, and part of it is batching that any MoE gets for free. That isn't the VKAE claim, and we'll present it as what it is rather than fold it into the speedup.
What the paper leads with is the controlled comparison: identical concurrency, batch, and harness, VKAE on vs. off. Single-stream and the throughput ceiling stay as separate, labeled context.
Same discipline on quality β the tail, not the mean: long-context recall and the hard subset, measured against VKAE-off on identical inputs. If it doesn't hold there, it doesn't ship.
To your question directly: VKAE-on vs. VKAE-off at identical concurrency. That's the headline.
reacted to their post with β€οΈ 3 days ago
Post
5080
π We ran genuine quantum key-recovery on 'real IBM quantum hardware' β and pushed the frontier well past the largest hardware demos we're aware of (which sat at N=4).
Using Simon's algorithm on
β’ EvenβMansour β N=5 β N=10
β’ 3-round Feistel (DES-family) β block 6 β 8
Each verified against an 'independent control key', using error mitigation only (no QEC).
π§ Honest scope: this is not a quantum speedup (the effective difficulty tracks the classical birthday bound ~2^{n/2}), not a break of real AES/RSA, and not 16-round DES (ours is 3-round). The recovery method is reserved for a forthcoming paper; formal record status is pending peer review.
π Write-up: https://huggingface.co/blog/FINAL-Bench/quantum
πΉοΈ Try it live in your browser: https://vidraft-quantumos.hf.space/crypto
π Leaderboard: FINAL-Bench/quantum-bench-leaderboard
#quantum #cryptography #quantumcomputing
Using Simon's algorithm on
ibm_kingston, we recovered the secret key of two symmetric-cipher structures:β’ EvenβMansour β N=5 β N=10
β’ 3-round Feistel (DES-family) β block 6 β 8
Each verified against an 'independent control key', using error mitigation only (no QEC).
π§ Honest scope: this is not a quantum speedup (the effective difficulty tracks the classical birthday bound ~2^{n/2}), not a break of real AES/RSA, and not 16-round DES (ours is 3-round). The recovery method is reserved for a forthcoming paper; formal record status is pending peer review.
π Write-up: https://huggingface.co/blog/FINAL-Bench/quantum
πΉοΈ Try it live in your browser: https://vidraft-quantumos.hf.space/crypto
π Leaderboard: FINAL-Bench/quantum-bench-leaderboard
#quantum #cryptography #quantumcomputing
fp8 of DARWIN 398b BF16?
π₯ 1
2
#1 opened 4 days ago
by 
hfmefi69
upvoted an article 4 days ago
Article
Quantum Cryptanalysis on Real Hardware: Pushing Symmetric-Structure Key Recovery Beyond the Published Frontier
FINAL-Bench
β’ β’ 15 reacted to their post with π₯ 4 days ago
Post
5080
π We ran genuine quantum key-recovery on 'real IBM quantum hardware' β and pushed the frontier well past the largest hardware demos we're aware of (which sat at N=4).
Using Simon's algorithm on
β’ EvenβMansour β N=5 β N=10
β’ 3-round Feistel (DES-family) β block 6 β 8
Each verified against an 'independent control key', using error mitigation only (no QEC).
π§ Honest scope: this is not a quantum speedup (the effective difficulty tracks the classical birthday bound ~2^{n/2}), not a break of real AES/RSA, and not 16-round DES (ours is 3-round). The recovery method is reserved for a forthcoming paper; formal record status is pending peer review.
π Write-up: https://huggingface.co/blog/FINAL-Bench/quantum
πΉοΈ Try it live in your browser: https://vidraft-quantumos.hf.space/crypto
π Leaderboard: FINAL-Bench/quantum-bench-leaderboard
#quantum #cryptography #quantumcomputing
Using Simon's algorithm on
ibm_kingston, we recovered the secret key of two symmetric-cipher structures:β’ EvenβMansour β N=5 β N=10
β’ 3-round Feistel (DES-family) β block 6 β 8
Each verified against an 'independent control key', using error mitigation only (no QEC).
π§ Honest scope: this is not a quantum speedup (the effective difficulty tracks the classical birthday bound ~2^{n/2}), not a break of real AES/RSA, and not 16-round DES (ours is 3-round). The recovery method is reserved for a forthcoming paper; formal record status is pending peer review.
π Write-up: https://huggingface.co/blog/FINAL-Bench/quantum
πΉοΈ Try it live in your browser: https://vidraft-quantumos.hf.space/crypto
π Leaderboard: FINAL-Bench/quantum-bench-leaderboard
#quantum #cryptography #quantumcomputing
posted an update 4 days ago
Post
5080
π We ran genuine quantum key-recovery on 'real IBM quantum hardware' β and pushed the frontier well past the largest hardware demos we're aware of (which sat at N=4).
Using Simon's algorithm on
β’ EvenβMansour β N=5 β N=10
β’ 3-round Feistel (DES-family) β block 6 β 8
Each verified against an 'independent control key', using error mitigation only (no QEC).
π§ Honest scope: this is not a quantum speedup (the effective difficulty tracks the classical birthday bound ~2^{n/2}), not a break of real AES/RSA, and not 16-round DES (ours is 3-round). The recovery method is reserved for a forthcoming paper; formal record status is pending peer review.
π Write-up: https://huggingface.co/blog/FINAL-Bench/quantum
πΉοΈ Try it live in your browser: https://vidraft-quantumos.hf.space/crypto
π Leaderboard: FINAL-Bench/quantum-bench-leaderboard
#quantum #cryptography #quantumcomputing
Using Simon's algorithm on
ibm_kingston, we recovered the secret key of two symmetric-cipher structures:β’ EvenβMansour β N=5 β N=10
β’ 3-round Feistel (DES-family) β block 6 β 8
Each verified against an 'independent control key', using error mitigation only (no QEC).
π§ Honest scope: this is not a quantum speedup (the effective difficulty tracks the classical birthday bound ~2^{n/2}), not a break of real AES/RSA, and not 16-round DES (ours is 3-round). The recovery method is reserved for a forthcoming paper; formal record status is pending peer review.
π Write-up: https://huggingface.co/blog/FINAL-Bench/quantum
πΉοΈ Try it live in your browser: https://vidraft-quantumos.hf.space/crypto
π Leaderboard: FINAL-Bench/quantum-bench-leaderboard
#quantum #cryptography #quantumcomputing
published an article 4 days ago
Article
Quantum Cryptanalysis on Real Hardware: Pushing Symmetric-Structure Key Recovery Beyond the Published Frontier
FINAL-Bench
β’ β’ 15 updated a
Space 4 days ago
updated 2
models 5 days ago
upvoted a collection 5 days ago
updated 2
collections 5 days ago
upvoted an article 6 days ago
Article
Adding a GPU Without Building One
FINAL-Bench
β’ β’ 17 reacted to their post with β€οΈ 6 days ago
Post
2989
π Adding a GPU without building one
AI is usually framed as "how smart is the model / how many GPUs did you buy." The real bottleneck is elsewhere β how efficiently you use the GPUs you already have.
Training happens once; inference runs the entire time users use your product. So a service's economics come down to cost per token. Inference acceleration uses software to pull several times more out of the same GPU β the effect of plugging in one more "virtual GPU."
VIDRAFT's VKAE, measured (B200, same-harness, no quality loss):
Qwen3.5-35B-A3B (MoE): 25.7 β 601 tok/s (23.4Γ)
Darwin-36B-Opus (in-house MoE): 25.0 β 280.8 (11.2Γ)
10,000+ tok/s peak aggregate under concurrency
The key: it's reproducible β model + serving shipped as one container.
docker pull vidraft/qwen35-vkae:601
Don't take our word for it β run it yourself. The mechanism will be released as a paper.
π Leaderboard & demo π VIDraft/vkae
Articles π https://huggingface.co/blog/FINAL-Bench/vkae-leaderboard
AI is usually framed as "how smart is the model / how many GPUs did you buy." The real bottleneck is elsewhere β how efficiently you use the GPUs you already have.
Training happens once; inference runs the entire time users use your product. So a service's economics come down to cost per token. Inference acceleration uses software to pull several times more out of the same GPU β the effect of plugging in one more "virtual GPU."
VIDRAFT's VKAE, measured (B200, same-harness, no quality loss):
Qwen3.5-35B-A3B (MoE): 25.7 β 601 tok/s (23.4Γ)
Darwin-36B-Opus (in-house MoE): 25.0 β 280.8 (11.2Γ)
10,000+ tok/s peak aggregate under concurrency
The key: it's reproducible β model + serving shipped as one container.
docker pull vidraft/qwen35-vkae:601
Don't take our word for it β run it yourself. The mechanism will be released as a paper.
π Leaderboard & demo π VIDraft/vkae
Articles π https://huggingface.co/blog/FINAL-Bench/vkae-leaderboard
- 3 replies
posted an update 6 days ago
Post
2989
π Adding a GPU without building one
AI is usually framed as "how smart is the model / how many GPUs did you buy." The real bottleneck is elsewhere β how efficiently you use the GPUs you already have.
Training happens once; inference runs the entire time users use your product. So a service's economics come down to cost per token. Inference acceleration uses software to pull several times more out of the same GPU β the effect of plugging in one more "virtual GPU."
VIDRAFT's VKAE, measured (B200, same-harness, no quality loss):
Qwen3.5-35B-A3B (MoE): 25.7 β 601 tok/s (23.4Γ)
Darwin-36B-Opus (in-house MoE): 25.0 β 280.8 (11.2Γ)
10,000+ tok/s peak aggregate under concurrency
The key: it's reproducible β model + serving shipped as one container.
docker pull vidraft/qwen35-vkae:601
Don't take our word for it β run it yourself. The mechanism will be released as a paper.
π Leaderboard & demo π VIDraft/vkae
Articles π https://huggingface.co/blog/FINAL-Bench/vkae-leaderboard
AI is usually framed as "how smart is the model / how many GPUs did you buy." The real bottleneck is elsewhere β how efficiently you use the GPUs you already have.
Training happens once; inference runs the entire time users use your product. So a service's economics come down to cost per token. Inference acceleration uses software to pull several times more out of the same GPU β the effect of plugging in one more "virtual GPU."
VIDRAFT's VKAE, measured (B200, same-harness, no quality loss):
Qwen3.5-35B-A3B (MoE): 25.7 β 601 tok/s (23.4Γ)
Darwin-36B-Opus (in-house MoE): 25.0 β 280.8 (11.2Γ)
10,000+ tok/s peak aggregate under concurrency
The key: it's reproducible β model + serving shipped as one container.
docker pull vidraft/qwen35-vkae:601
Don't take our word for it β run it yourself. The mechanism will be released as a paper.
π Leaderboard & demo π VIDraft/vkae
Articles π https://huggingface.co/blog/FINAL-Bench/vkae-leaderboard
- 3 replies
published an article 6 days ago
Article
Adding a GPU Without Building One
FINAL-Bench
β’ β’ 17