CNRS NLP researcher working on multilingual counterspeech against hate speech and LLMs for critical science education.
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-15) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.
Research profile
Output—
Not yet scored: this snapshot has no aggregate publication counts, venue profile, or authorship-role data, so research output cannot be quantified here. This axis would measure the volume and venues of papers over time, including first and corresponding authorship.
Impact—
Not yet scored: no citation counts, h-index, or influence indicators are available in this snapshot. This axis would measure the citation-based uptake of the two cataloged papers across the counterspeech and science-education NLP communities over time.
Focus0/100
The focus score is 0, and the raw field distribution explains why: the profile splits evenly, with 0.5 on natural language processing for hate speech / counterspeech generation and 0.5 on NLP for science education applications. The two lines share methodological vocabulary — stereotypes, implied statements, annotation schemas, argument mining — but spanning two distinct application areas keeps measured focus low in this snapshot.
Collaboration55/100
Collaboration scores 55, built on 15 distinct coauthors in the snapshot. The recurring list — Serena Villata, Greta Damo, Elias Urios Alacreu, Paolo Rosso, Rodrigo Agerri, Itziar Aldabe, Mark Cieliebak, Jan Deriu, Mariana Flores, Jurgita Kapociute-Dzikiene, Dovile Kuiziniene, Arantza Rico — points to steady multi-author teamwork spanning both the counterspeech and science-education lines rather than solitary work.
Activity—
Not yet scored: per-year publication counts and recency measures are absent, so research tempo cannot be assessed. This axis would track output cadence over time; both cataloged papers are dated 2026, which alone cannot establish an activity trend.
Reproducibility38/100
Reproducibility averages 38 across the 2 papers assessed. The counterspeech paper is rated medium readiness — evidence for stereotype-conditioned prompting is strong, but inference depends on gold annotations — while the CRITICS paper is rated low readiness as a proposal without results. Reproducing this line of work requires released annotated data and multilingual evaluation setups.
Research summary
Elena Cabrio is a researcher at the Centre National de la Recherche Scientifique (CNRS); this profile is compiled from ORCID-linked bibliographic records (ORCID 0000-0001-9374-7872) and papers reviewed on Evolve, and should be read as an early-stage, evidence-bounded view. Her visible recent work sits in two natural language processing lines that split the field distribution evenly (0.5 each): hate speech and counterspeech generation, and NLP for science education. The 2026 arXiv paper 'Deconstructing Stereotypes' proposes scope-conditioned generation for multilingual counterspeech; Evolve's review finds stereotype-conditioned prompting strongly supported while flagging reliance on gold annotations at inference time (medium reproducibility readiness). The companion 2026 output, 'CRITICS — Critical Science Without Borders', sets out a project uniting LLM-based scientific machine translation — including low-resource languages such as Basque — with critical-thinking education; it is rated a well-grounded but results-free proposal (low readiness). Collaboration data records 15 distinct coauthors (score 55), with Serena Villata, Greta Damo, Paolo Rosso, Rodrigo Agerri, Itziar Aldabe and others recurring. Reproducibility averages 38 across 2 assessed papers, while output, citation impact, and activity metrics are not yet available.
Collaboration pattern
With 15 distinct coauthors and a collaboration score of 55, the pattern in this snapshot is steady multi-author teamwork. The recurring list centers on Serena Villata and Greta Damo and extends to Elias Urios Alacreu, Paolo Rosso, Rodrigo Agerri, Itziar Aldabe, Mark Cieliebak, Jan Deriu, Mariana Flores, Jurgita Kapociute-Dzikiene, Dovile Kuiziniene, and Arantza Rico, spanning both the counterspeech and science-education lines.
Research keywords
Highlights
Scope-conditioned multilingual counterspeech generation
CRITICS: language models for critical thinking in science education
Annotation and argumentation as the connective thread
Across both cataloged lines, annotation schemas, implied statements, and argument mining recur as shared methodological vocabulary, linking counterspeech generation to critical-thinking education within one NLP agenda.
Timeline
- 2026Posts 'Deconstructing Stereotypes: Scope-Conditioned Generation for Effective Multilingual Counterspeech' to arXiv.
- 2026Releases the CRITICS project paper on language models to promote critical thinking in science education.
- 2026Both cataloged works appear in the same year, establishing parallel lines in counterspeech generation and science-education NLP.
Papers in the Evolve catalog
Frequent co-authors
About this profile
This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.
- Bibliometric signals come from a limited snapshot; the output, impact, and activity dimensions lack underlying data and are left unscored rather than estimated.
- Only 2 papers — both 2026 arXiv preprints — are reviewed in the Evolve catalog; earlier or peer-reviewed work is not represented in this profile.
- Reproducibility figures (mean 38) cover just 2 assessed papers and should not be generalized to the full portfolio.
- Country-level affiliation is missing in the source record (institution given as CNRS only); the profile is name-resolved from public bibliographic data via ORCID 0000-0001-9374-7872.