INDEX

Explanations

sections related to introductory information and prefaces in a textual document

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

wat

-0.73

wiki

-0.70

 pasture

-0.66

 unpredictable

-0.65

 caster

-0.65

 eyed

-0.65

ãĥ¯ãĥ³

-0.65

ousel

-0.63

fly

-0.61

 lookout

-0.60

POSITIVE LOGITS

 Letters

0.84

 thanking

0.83

 aloud

0.82

aug

0.82

 thereto

0.82

enza

0.78

Letter

0.77

 Letter

0.74

matter

0.72

notes

0.72

Activations Density 15.197%

sections related to introductory information and prefaces in a textual document

No Comments

No Known Activations