INDEX

Explanations

proper nouns related to military, politics, and international events

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

hire

-1.02

lez

-0.86

etr

-0.85

ki

-0.76

icular

-0.74

engers

-0.74

ements

-0.72

hement

-0.71

wikipedia

-0.71

eways

-0.69

POSITIVE LOGITS

ufact

0.86

otive

0.86

 Bates

0.74

essage

0.73

 Coleman

0.71

 Reed

0.68

oute

0.66

FG

0.66

ONEY

0.64

parts

0.64

Activations Density 8.540%

proper nouns related to military, politics, and international events

No Comments

No Known Activations