INDEX

Explanations

full names of reporters

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

apollo-research/Skylion007-openwebtext-tokenizer-gpt2

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

perm

-0.80

inence

-0.75

ensions

-0.73

ertodd

-0.72

ateral

-0.72

emonic

-0.72

uras

-0.68

liga

-0.67

ersen

-0.66

ujah

-0.65

POSITIVE LOGITS

 reporter

0.78

 corps

0.77

 Seymour

0.77

 covering

0.76

ially

0.74

ials

0.72

 COVER

0.69

 reporters

0.69

 interviewed

0.67

 Coverage

0.67

Activations Density 7.380%

full names of reporters

No Comments

No Known Activations