Claim Review Queue
Review scaffolded ClaimCards from arXiv and manual submissions. Approve to add to leaderboard at Draft tier.
How does this page work?
What is a ClaimCard? A ClaimCard is a structured record of a scientific result claim — typically from a paper. It captures who claimed the result, what they claimed (method, PSNR, SSIM), and where it was published.
What does "Scaffold" mean? Scaffolding creates a new ClaimCard and puts it in the review queue at "Pending" status. It does NOT publish the result to the leaderboard — a reviewer must approve it first.
What does "Approve" do? Approving moves the claim to the benchmark leaderboard at Draft trust tier. The result is now visible but not yet verified.
What does "Reject" do? Rejecting removes the claim from the queue with a reason. The submitter may re-apply with corrections.
What does "Mark Reproduced" do? When a Benchmark Reviewer independently confirms the result, they click this to upgrade the trust tier from Draft to Reproduced.
What does "Demote to Draft" do? If a reproduction fails or an error is found, a reviewer can demote the claim back to Draft for re-evaluation.
Trust-tier progression:
Draft → Author-confirmed → Reproduced → Certified
Example 1: Single scaffold (one paper)
1. Click "+ Scaffold new ClaimCard" below
2. Fill in:
| arXiv ID: | 2204.07908 |
| Title: | MST: Mask-guided Spectral-wise Transformer |
| Authors: | Yuanhao Cai, Jing Lin, Haoqian Wang |
| Modality: | cassi |
| Method: | MST |
| PSNR: | 35.18 |
| SSIM: | 0.948 |
3. Click "Scaffold ClaimCard" → claim appears in queue with "Pending" badge
4. Click "Approve" → claim goes to CASSI leaderboard at Draft tier
Example 2: Batch scaffold (multiple papers at once)
1. Click "+ Batch scaffold from arXiv IDs" below
2. Enter multiple arXiv IDs, one per line:
2204.07908 2305.18727 2312.00001
3. Click "Scaffold Batch" → the system fetches each paper from arXiv, scores relevance, and creates ClaimCards for relevant ones
4. Claims appear in the queue below — review and approve each one individually
Note: papers not related to computational imaging (relevance < 0.2) are automatically filtered out.
+ Scaffold new ClaimCard (create a new paper claim for review)
+ Batch scaffold from arXiv IDs
DiffusionCT: Latent Diffusion Model for CT Image Standardization
arxiv:2305.18727 · Md Selim, Jie Zhang, Michael A. Brooks · ct · DiffusionCT
Claimed: PSNR=38.2 dB SSIM=0.965