INDEX

Explanations

names, especially those with variations in punctuation

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

kefeller

-1.00

acca

-0.82

icum

-0.80

iqueness

-0.71

Interstitial

-0.68

ities

-0.64

iets

-0.63

ishment

-0.61

ideo

-0.61

unity

-0.60

POSITIVE LOGITS

hog

0.99

orge

0.85

ORGE

0.84

rer

0.80

rers

0.74

eks

0.71

ois

0.67

cko

0.65

aux

0.65

orget

0.65

Activations Density 11.361%

names, especially those with variations in punctuation

No Comments

No Known Activations