INDEX

Explanations

phrases that express contradictions or shortcomings

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Embeds

IFrame

Link

Negative Logits

 conform

-0.06

Ð³Ñĥ

-0.06

 columnIndex

-0.06

 foll

-0.05

insky

-0.05

Ø§ØµØ±

-0.05

Explicit

-0.05

Cul

-0.05

 Lynch

-0.05

.annotate

-0.05

POSITIVE LOGITS

 statement

0.11

statement

0.10

 Statement

0.10

 untrue

0.09

Statement

0.09

 Actually

0.08

 statements

0.08

Actually

0.08

 actually

0.08

_statement

0.07

Activations Density 0.029%

No Known Activations