INDEX

Explanations

occurrences of specific place names

oai_token-act-pair · gpt-3.5-turbo

references to the location "Westminster."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Hulk

-0.75

 analy

-0.71

roid

-0.70

 infected

-0.67

>[

-0.67

izer

-0.66

 activ

-0.65

iment

-0.64

 favors

-0.63

eria

-0.61

POSITIVE LOGITS

 Westminster

4.07

 Downing

1.70

 Edinburgh

1.62

minster

1.56

 Surrey

1.45

Scotland

1.41

London

1.39

 Notting

1.37

 Glasgow

1.35

 Buckingham

1.32

Activations Density 0.017%

occurrences of specific place names

references to the location "Westminster."

No Comments

No Known Activations