Direction, evidence grade, and study type are AI-generated labels (gpt-5-mini), not human-verified. Syntheses are LLM-written. "Tensions" are machine-detected candidates, not confirmed contradictions. A research-acceleration tool, not peer review.
How this is built →
2Distinct papers
7Unique collaborators
2/2Semantic Scholar citation coverage
Publication span: 2026. Corpus fetch span: 2026.
Identity provenance
Provider IDs
- Semantic Scholar:
2302152737
ORCID evidence
No valid ORCID is stored.
Observed aliases (1)
- Benjamin Zelditch (semantic scholar, provider refresh)
Topics and outcomes in this view
Assessment themes
- Productivity: 2 papers
- Adoption: 1 paper
- Innovation: 1 paper
Claim outcomes
- Firm Revenue: 2 papers
- Organizational Efficiency: 1 paper
- Other: 1 paper
- Automation Exposure: 1 paper
- Decision Quality: 1 paper
- Firm Productivity: 1 paper
- Innovation Output: 1 paper
Papers in the Semantic Scholar view
Latest stored Semantic Scholar author observations only. Citation counts below are from the same provider and are not combined with other services.
Scroll the table horizontally to see every column.
| Paper | Author evidence | Date | Provider citations |
|---|---|---|---|
| A production causal optimization pipeline that explicitly targets incremental impact—combining uplift neural nets, uncertainty-aware bandit exploration, and large-scale constrained allocation—raised LinkedIn Feed marketing long-term value by 7.2% in an online A/B test (p=0.041), showing that production-scale causal targeting under business constraints is feasible.arxiv | Benjamin Zelditch provider id |
2026-08-10 | 0 |
| BanditLP pairs neural Thompson Sampling with a billion-variable LP solver to deliver constrained contextual bandits at web scale, improving performance on benchmarks and driving a documented business uplift in LinkedIn's email marketing.arxiv | Benjamin Zelditch provider id |
2026-01-22 | 1 |
Citation observation summary
Semantic Scholar supplied counts for 2 of 2 papers in this view; 0 are missing. The observed paper counts sum to 1 cumulative citations. This is a coverage summary, not an author score or h-index.