LitPivot develops research ideas by dynamically contextualizing them within the literature landscape.

Meta description: LitPivot helps researchers revise early ideas using literature; a study with 17 researchers found better grounding, novelty and usefulness in proposals.

Tags: LitPivot, Asta, research ideation, UIST 2026, Ai2

Featured image: abstract, friendly illustration of linked paper cards and a highlighted proposal panel, with no real people, no logos and no text.

LitPivot helps researchers sharpen early research proposals by checking each part of a draft against prior literature. Developed by researchers at Ai2 with collaborators at the University of Pennsylvania, it uses papers retrieved through Asta Paper Finder, groups them around the proposal segment being edited, offers critiques supported by excerpts from selected studies, and suggests revisions that users can accept, edit or reject. In a study with 17 researchers, two blind domain experts rated LitPivot-assisted proposals as better grounded in literature than proposals made with a text editor and an AI chat tool; participants also reported stronger literature understanding and greater perceived novelty and usefulness. The work is scheduled for presentation at UIST 2026 in Detroit.

Why Literature Can Force a Pivot

A new paper can change a research idea's trajectory: it may show that a proposed contribution has already been addressed, or expose an unresolved problem that opens a new direction. The team calls this a literature-initiated pivot, linking two activities researchers often manage separately: developing an idea and choosing the literature to measure it against.

An analysis of meeting notes from four completed computer science research projects identified 21 literature-initiated pivots across those projects. They fell into two types: differentiating an idea from existing work to establish novelty, and aligning an idea with existing work to build on it. The LitPivot project itself went through two pivots before settling on its contributions.

In a formative study, five researchers developed proposals with a text editor and standard search tools. They struggled to develop ideas while critically evaluating them against the literature. When a paper challenged an idea's novelty, they wanted concrete revision suggestions supported by prior work. LitPivot's assessments pair literature-supported critiques with specific suggestions for revising the selected part of an idea.

How LitPivot Turns Papers Into Revision Prompts

In LitPivot, a researcher writes a short proposal in an editor and enables facet highlighting, which segments an idea into a problem, contribution and evaluation plan. Clicking a proposed facet opens an adjacent panel that groups related papers based on that facet; for an evaluation facet, papers may be grouped by the evaluation methods they use, with the most relevant groups highlighted. The researcher selects papers, requests an assessment, reviews a suggested revision alongside supporting excerpts, and the literature is reranked for relevance to the revised facet.

Each facet has its own literature set and relevance ranking. LitPivot groups retrieved papers around the facet being edited, letting the researcher choose which ones the system should use to assess the draft. For a problem statement, it looks for evidence that the problem exists and reasons it matters. For a proposed contribution, it compares what selected papers have already addressed with their limitations and future work to look for remaining gaps. For an evaluation plan, it checks whether proposed tests address the research question and how similar work has been evaluated. The system then suggests a rewrite with supporting quotes.

Changing one facet can make another out of date. LitPivot flags passages that may need revision, such as an evaluation plan that no longer adequately evaluates a revised contribution. Researchers can accept, edit or reject suggestions, request new papers or refresh the literature ranking as the proposal develops.

Alongside scientific-writing environments such as Prism and agentic assistants such as Claude Code, LitPivot explores a new workflow for developing an early research idea: choosing literature to assess it against, using that evidence to revise it, and reconsidering which papers matter as the idea changes. Its paper groups and linked critiques make those steps explicit for each part of the proposal, while researchers control both the draft and the evidence used to assess it.

What the 17-Researcher Study Found

In the lab study, 17 researchers each developed one assigned idea with LitPivot and another with a baseline tool: a text editor and an AI tool for asking questions about selected papers. Both conditions used the same collection of papers. Two domain experts, unaware of which system had been used, rated the resulting proposals on a single criterion: how well they used relevant literature to support and justify the idea. LitPivot proposals received higher ratings.

Participants also reported a stronger understanding of the surrounding literature and greater self-reported improvements in novelty and usefulness with LitPivot than with the baseline. Domain experts reviewed the participants' ideas and rated ideas revised with LitPivot as significantly better grounded in relevant literature.

One participant began with a proposal to generate consistent, factual synthetic medical notes. LitPivot identified consistency across multiple hospital visits as a potential gap in the selected literature. The participant reviewed abstracts and removed papers about single-visit notes. After another assessment, the proposal narrowed to detecting contradictions across notes from multiple visits. The example shows how revising an idea and choosing the literature used to assess it can inform each other.

From Early Ideas to Ongoing Projects

In a follow-up study, five researchers used LitPivot on their own ongoing projects during 30-minute sessions. They used it across multiple project phases, not just at the beginning. Two used it to vet new ideas against the literature, while three used it to refine the framing of more mature ideas. One researcher highlighted how seeing parallel concepts from different domains helped reframe their contribution.

These results suggest LitPivot can support early exploration and refinement of an existing research direction. It is intended for early-stage research ideation, but could extend to longer writing tasks such as writing a paper introduction.

At Ai2, the broader effort aims to help researchers formulate testable ideas and examine the evidence behind them. Asta supports that effort through theory generation from published findings with Theorizer, synthesis, and hypothesis generation and testing on data with AutoDiscovery. HARPA generates promising hypotheses informed by emerging research trends. Scideator lets researchers recombine elements of prior work into new ideas and assess novelty, and IdeaSynth supports iterative refinement with feedback grounded in literature.

LitPivot extends this line of work. A recent study, Old Ideas, Novel Problems, examined limits of automated novelty evaluation and showed how small prompt changes can alter an AI judge's verdict on the same ideas. LitPivot makes the choice of relevant literature part of the revision process: researchers can change which papers assess a proposal as the idea develops, use critiques and supporting excerpts to decide how to revise, and connect those decisions across the problem, contribution and evaluation plan. The goal is to help scientists build on existing knowledge and turn early ideas into new contributions to their fields.

Who Should Use It / Who Should Skip It

Researchers who need to test a draft idea against prior work, especially early-stage proposals with problem, contribution and evaluation parts, may benefit most. Those who only need final writing support may wait, because LitPivot is intended for early-stage ideation, though it could extend to introduction sections. Because it relies on selected papers and literature ranking, users who want fully automated novelty judgments should treat its suggestions as reviewable, not final.