LawGrader
Help Center
Grading and feedback

When a grade needs a second try

Stopping a run, re-grading a paper, what a failed or incomplete grade means, and what happens during an AI outage.

Updated September 12, 2026

Most grades finish in a minute or two and never need attention. This article is for the other cases: a run you want to stop, a paper you want graded again, a run that failed, and the days when an AI provider is down.

Grading a paper, and stopping it

The control at the top of the grading page is the assist dial. On an ungraded paper its button reads Grade this paper; press it and the paper goes to the assistant. While it runs the button becomes Stop grading: stop it and nothing is saved and no credit is spent. A stopped paper goes back to Ready to grade.

Re-grading

On a graded paper the same control offers Re-grade this paper. It asks first, because every score is replaced, including any you set by hand, and so are the AI's comments. Comments you wrote or edited stay. Re-grading never costs a second credit. The two reasons to do it:

  • The rubric changed since the paper was graded. A strip across the top of the page says so, names what changed, and carries its own Re-grade button; re-grading refreshes the scores against the current rubric.
  • You changed grading guidance, approved a learned rule from validation, or added supporting material and want this paper read under it.

On the assignment page, Re-grade all does the same for every graded paper at once; see Grading a whole class at once. A curved paper that is re-graded rejoins the curve at its new score.

Failed

Failed in the paper list means the run did not finish. Open the paper and the dial's message says which of two things happened: the run failed outright and the paper is unchanged, or it failed partway and the scores that finished are on the rubric with annotations possibly missing. Either way, run it again. A transient error (a timeout, a provider hiccup) is retried on its own before a paper is ever marked failed, so a Failed you see is usually worth a look at the file: a corrupt PDF or an empty document fails the same way twice.

A paper stranded mid-grade by a deploy or a crash is picked up again automatically; you do not need to do anything.

Incomplete

Incomplete in the score column is different from Failed. The run finished, but some criteria came back unscored, most often because part of the paper could not be read for a criterion judged from the look of the page. Those criteria are left for you rather than guessed at. The paper will not release with an Incomplete grade; open it, score what is left, and confirm. See How grading works.

During an outage

When an AI provider is unreachable, an amber notice appears at the top of the app: grading may take longer, and your work is safe. Papers already queued show Waiting for the grader instead of Queued, and they run as soon as service returns.

If the main grader is down and you press Grade anyway, you get a choice:

  • Grade now with the backup model. Results right away from a different provider, which may score a little differently, so you may want to spot-check them. A paper graded this way is marked in the list ("Graded by the backup model") and is held for your review rather than confirmed automatically. Re-grade it later to use the main grader.
  • Schedule for when the main grader is back. Same-quality results, later, automatically.

Either way nothing is lost and nothing is charged twice.

Was this helpful?

Keep reading

  • Uploading papers — Which files work, how a stack is matched to students, and what each status in the paper list means.
  • How grading works — The assistant proposes; you decide. How scores are made and how you stay in control.
  • Scoring a criterion with rules — When a criterion is a checklist rather than a judgment: Binary and Count rules, adding up or deducting, and what the rail shows.
  • Validating the grader against your grading — Run a few papers you have already graded on the Validate step, so the assistant learns how you actually score before it grades the class.
  • Grading a whole class at once — Once validation satisfies you, grade the whole set in one pass, then review the results.
  • Reviewing and adjusting the feedback — Adopt the assistant's observations you agree with, dismiss the rest, and write your own.
  • Reading the grade rail — How much of the AI's read to show, where each score came from, what the look-fors say, and how to filter the margin.
  • Marking up a paper — The grading page: the pen tools, your own comments, suggested rewrites, what students see, and undo.
  • The assistant, and talking to it — The three places you can ask the assistant for help, what each one knows, and how to dictate instead of type.
  • Your comment library — Every comment you write collects in one library: browse it while grading, insert with a keystroke, and bring in the comments you wrote before.
  • Comment suggestions in your own voice — As you write a margin comment, LawGrader suggests the rest of it, drawn from the comments you have written before.
  • Giving the grader supporting docs — Hand the assistant the documents you grade against: the set reading, a reference text, a style guide.