INDEX

Explanations

instances of the word "however."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_slefr-ajt/2-res_slefr-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.2.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

uto

-1.00

foundation

-0.93

enders

-0.89

amide

-0.88

 Architects

-0.85

iris

-0.84

anders

-0.84

holders

-0.81

 Â£

-0.80

 Bigfoot

-0.79

POSITIVE LOGITS

srf

1.19

CLASSIFIED

1.13

wcs

1.06

ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ

0.99

 unfocusedRange

0.99

opez

0.98

 guiActiveUn

0.95

 reluct

0.94

 exha

0.94

 arrang

0.94

Activations Density 0.094%

instances of the word "however."

No Comments

No Known Activations