INDEX

Explanations

names containing the string "Ed"

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Ed"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 redistributed

-0.66

 bluff

-0.66

 sight

-0.63

 retribution

-0.61

 theaters

-0.59

 shotguns

-0.57

 resulting

-0.57

 true

-0.57

 sensitive

-0.56

 favour

-0.56

POSITIVE LOGITS

monds

1.37

gew

1.30

monton

1.29

mond

1.26

dy

1.25

elman

1.23

mund

1.17

ifice

1.16

itions

1.12

gewater

1.11

Activations Density 0.021%

names containing the string "Ed"

mentions of the name "Ed"

No Comments

No Known Activations