Know what
reviewers will say
before they say it.

PaperMate runs your manuscript through an agentic reviewer — surfacing weaknesses, identifying unsupported claims, and scoring across all 7 ACL ARR metrics before submission day.

Review my paper See how it works
ARR
Rubric Standard
<10m
Avg. Review Time
7
ARR Metrics
How it works

From PDF to peer review
in three steps

Upload once, get the kind of structured feedback that usually takes weeks to receive from a conference.

1

Upload your manuscript

Sign in, drop your PDF, and confirm your submission meets ACL ARR requirements. Your secret key is generated instantly.

2

AI reviews your paper

Our agentic reviewer analyzes your paper against the ACL ARR rubric, cross-references the latest literature, and generates a structured evaluation — all in minutes.

3

Read your review

Enter your secret key to access a full structured review: all 7 ACL ARR metrics, concrete strengths and weaknesses with citations, and actionable suggestions before you submit.

What you receive

A review that reads
like it's from a senior reviewer

PaperMate applies the ACL Rolling Review rubric — the same framework top-tier NLP conferences use — to give you feedback that's specific, calibrated, and actionable.

  • Paper summary in the reviewer's own words
  • 3–5 strengths & weaknesses with evidence
  • Comments, suggestions & ethics assessment
  • 7 ARR metrics: Soundness, Excitement, Reproducibility, Datasets, Software, Confidence + Overall Assessment
Get my review
papermate — review
Overall Assessment
Borderline Accept — Conference
3.5 / 4
Soundness
4/5
Excitement
3/5
Repro.
4/5
Datasets
5/5
Software
4/5
Confidence
4/5
Strengths
Weaknesses
Demo

See PaperMate in action

Watch how PaperMate reviews a real NLP paper — from PDF upload to a full ACL ARR-style evaluation.

The rubric

Every score explained

PaperMate follows the ACL Rolling Review rubric precisely — the same criteria top NLP conferences use to evaluate submissions.

S

Soundness

Scientific and technical rigor. Are the methodology and experimental design valid? Do claims have sufficient empirical support?

E

Excitement

Impact and novelty. Is this a breakthrough, a solid contribution, or incremental work? Will it influence future research directions?

R

Reproducibility

Can the community replicate the results? Code release, hyperparameter reporting, and implementation detail all factor in.

D

Datasets

How much value does any released dataset add to the community — from enabling entirely new research directions to no release at all.

SW

Software

How much value does any released code or tooling add — from enabling others' research to no release at all.

C

Confidence

Reviewer's self-assessed domain expertise. Calibrates how much weight meta-reviewers should place on this evaluation.

OA

Overall Assessment

The bottom-line recommendation — whether the paper should be rejected, accepted to Findings, accepted to the main conference, or considered for an award. Mirrors the ACL ARR scale exactly.

Ready to know what
reviewers will say?

Sign in, upload your PDF, and get a structured peer review
in under 10 minutes.

Submit my paper now View existing review