INDEX

Explanations

mentions of specific geographical locations, specifically towns

oai_token-act-pair · gpt-3.5-turbo

mentions of a specific town or locality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

aunder

-0.92

utic

-0.91

uberty

-0.85

utics

-0.83

emis

-0.80

etheus

-0.75

ĻĤ

-0.75

 nipple

-0.73

ollo

-0.73

ances

-0.71

POSITIVE LOGITS

spe

1.18

folk

1.09

ships

1.05

 Halls

1.03

houses

0.97

scape

0.97

house

0.89

side

0.83

ship

0.83

 Town

0.82

Activations Density 0.011%

mentions of specific geographical locations, specifically towns

mentions of a specific town or locality

No Comments

No Known Activations