Yash More
International Institute of Information Technology, Hyderabad
IIIT Hyderabad NLP researcher spanning cost-effective machine translation and LLM-based peer-review generation.
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—
The current snapshot provides no publication counts, venue list, or authorship-role totals, so output is not scored. This axis would measure career paper volume, venue distribution, and first/corresponding-author share across the full record.
Impact—
No citation counts, h-index, or citing-work data are available, so impact is not scored. This axis would measure citation volume, influence of the two reviewed 2026 papers, and downstream adoption once citation data exists.
Focus0/100
The focus score reads 0 in this snapshot, but the raw field distribution behind it is narrow: 50% natural language processing and 50% natural language processing — machine translation, a single field family. Both reviewed 2026 papers fall inside that span, one on peer-review generation and one on machine translation, so the observed work concentrates on applied large-language-model studies rather than spreading across unrelated fields.
Collaboration67/100
Collaboration scores 67 on the strength of 28 distinct coauthors in the reviewed record, with 12 recurring partners. Two clusters stand out: an IIIT Hyderabad academic group including Manish Shrivastava, Vasudeva Varma and Maitreya Prafulla Chitale, and an industry machine-translation group around Tom Kocmi, Alexandre Bérard, Phil Blunsom and Nicholas Frosst on the Cohere CAT+ paper. On both catalog entries the name sits in the middle author list rather than as corresponding author.
Activity—
Per-year publication cadence cannot be computed because the snapshot supplies no annual output counts; only the two 2026 papers are in scope. This axis would measure yearly publication rate and recent momentum across the career.
Reproducibility55/100
Reproducibility averages 55 over the 2 assessed papers, both rated medium readiness. The consolidation study is a controlled Llama-3.1-8B experiment on MERC-36K, but Evolve warns the SEA-S ground truth may bias the comparison; the Cohere CAT+ translation paper describes training (difficulty-sampled SFT, DPO, online RL) transparently and evaluates honestly, yet its baselines are self-judged and lack significance tests, capping full replication confidence.
Research summary
Yash More is a natural language processing researcher affiliated with the International Institute of Information Technology, Hyderabad, whose reviewed work spans two applied strands of large-language-model research: machine translation and peer-review generation. The Evolve catalog holds two 2026 arXiv papers, in both of which Yash More appears as a middle author rather than corresponding author. The first, 'To Consolidate or not to Consolidate?', runs a controlled Llama-3.1-8B study on MERC-36K asking whether consolidating multiple references into one improves multi-reference training for peer-review generation; Evolve judges the design controlled but flags possible bias in the SEA-S ground truth. The second, 'North Small Translate' (Cohere CAT+), contributes to a claimed top-performing sub-1T machine translation system across 50 languages built with difficulty-sampled SFT, DPO and online reinforcement learning, praised for honest evaluation but relying on self-judged baselines. The collaboration record lists 28 distinct coauthors, splitting into an IIIT Hyderabad academic cluster around Manish Shrivastava and Vasudeva Varma and an industry machine-translation cluster around Tom Kocmi and Alexandre Bérard. Both papers rate medium reproducibility readiness, with a mean of 55.
Collaboration pattern
With 28 distinct coauthors and 12 recurring partners, Yash More works in a bridge position between an IIIT Hyderabad NLP group — Manish Shrivastava, Vasudeva Varma, Maitreya Prafulla Chitale, Harshit Gupta, Ketaki Mangesh Shetye, Manav Chaudhary — and industry machine-translation collaborators including Tom Kocmi, Alexandre Bérard, Phil Blunsom, Samuel Cahyawijaya, Shaun Cassini and Nicholas Frosst. On both reviewed 2026 papers Yash More appears as a middle author, indicating contribution within larger team efforts rather than independently led projects.
Research keywords
Highlights
Controlled test of reference consolidation in peer-review training
Cost-effective machine translation with Cohere CAT+
Timeline
- 2026Co-authored 'To Consolidate or not to Consolidate?', a controlled Llama-3.1-8B study of consolidation in multi-reference training for peer-review generation (arXiv, middle author).
- 2026Co-authored 'North Small Translate: Advanced Cost-Effective Translation (Cohere CAT+)', on difficulty-sampled SFT/DPO/online RL for sub-1T machine translation across 50 languages (arXiv, middle author).
- 2026Both papers entered the Evolve catalog with medium reproducibility readiness, giving a reproducibility mean of 55 across the 2 assessed papers.
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 dimension scores come from the platform's bibliometric snapshot; output, impact and activity axes are unscored because the underlying count and citation data were not available.
- Only 2 papers, both 2026 arXiv preprints, are in the reviewed catalog; the profile may not reflect the researcher's complete publication record.
- No country or ORCID is provided; “Yash More” is a relatively common name, so this profile is name-resolved from public bibliographic data and may carry mismatch risk.
- Reproducibility rests on only 2 assessed papers (mean 55, both medium readiness), too small a sample for a stable score.