INDEX

Explanations

words related to regions, medical conditions, and physical evidence

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

wcs

-0.86

lished

-0.74

ulhu

-0.73

emporary

-0.72

å§

-0.71

merce

-0.69

 Entered

-0.64

nown

-0.63

inctions

-0.61

apult

-0.61

POSITIVE LOGITS

ãĥ¼ãĥ

0.63

 Strauss

0.58

omaly

0.57

ãĥĳ

0.56

XM

0.56

heim

0.56

Ly

0.55

istant

0.55

versible

0.55

 Johann

0.55

Activations Density 12.855%

words related to regions, medical conditions, and physical evidence

No Comments

No Known Activations