Correlated errors: open-access research

Journal article

Perspectives Without Independence: Multi-Agent and Multi-Persona Reasoning Under Compute-Normalised Comparison, Why the Gains That Survive Are Not the Ones Diversity Predicts, and the Controls That Would Tell Them Apart

Pranay Mahendrakar 2026

Multi-agent debate, multi-persona prompting and related schemes are usually justified by diversity of viewpoint: several perspectives err differently, so their combination is more reliable than any one of them. That justification is a claim about independent errors, and a scheme that samples all…

Cite Related research LOR-00074 · 10 views · 22 downloads

PDF537.5 KB

Journal article

Consensus Too Soon, or Agreement From the Start? Shared Prior, Social Coupling and Pool Coverage in Decentralised LLM Collectives, and Why Prompted Diversity and Model Heterogeneity Act on Different Terms

Pranay Mahendrakar 2026

Groups of language-model agents that exchange answers and settle on a common one are now a standard way to build decentralised decision systems. A common worry is that they agree too soon: agents copy each other, diversity collapses, and the group loses the independent errors that make collective…

Cite Related research LOR-00057 · 6 views · 8 downloads

PDF14.6 MB

Journal article

A Second Model Is Not a Second Opinion: Peer Verification, Correlated Failure, and the Independence Variable Nobody Reports

Pranay M. Mahendrakar 2026

When a language model cannot reliably detect its own errors, the standard remedy is to have a second model check the first. That substitution is now the default across agent benchmarks, reward pipelines and safety evaluations, and it rests on an assumption that is rarely stated and that, in the…

Cite Related research LOR-00073 · 11 views · 38 downloads

PDF456.6 KB

Journal article

A Second Model Is Not a Second Opinion: Peer Verification, Correlated Failure, and the Independence Variable Nobody Reports

Pranay M. Mahendrakar 2026

When a language model cannot reliably detect its own errors, the standard remedy is to have a second model check the first. That substitution is now the default across agent benchmarks, reward pipelines and safety evaluations, and it rests on an assumption that is rarely stated and that, in the…

Cite Related research LOR-00043 · 7 views · 7 downloads

PDF451.8 KB

4 records

Perspectives Without Independence: Multi-Agent and Multi-Persona Reasoning Under Compute-Normalised Comparison, Why the Gains That Survive Are Not the Ones Diversity Predicts, and the Controls That Would Tell Them Apart

Consensus Too Soon, or Agreement From the Start? Shared Prior, Social Coupling and Pool Coverage in Decentralised LLM Collectives, and Why Prompted Diversity and Model Heterogeneity Act on Different Terms

A Second Model Is Not a Second Opinion: Peer Verification, Correlated Failure, and the Independence Variable Nobody Reports

A Second Model Is Not a Second Opinion: Peer Verification, Correlated Failure, and the Independence Variable Nobody Reports