INDEX

Explanations

phrases related to political statements or actions

oai_token-act-pair · gpt-3.5-turbo

commas and phrases indicating shifts in the narrative or context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ãĤ£

-0.80

ISE

-0.76

aimon

-0.76

ource

-0.75

ãĥīãĥ©

-0.73

aea

-0.70

aughtered

-0.70

dden

-0.69

itaire

-0.69

oleon

-0.67

POSITIVE LOGITS

 saying

1.63

 insisting

1.54

 telling

1.45

 tweeting

1.40

 stressing

1.40

 stating

1.38

 calling

1.38

 noting

1.37

 arguing

1.36

 dismissing

1.34

Activations Density 0.186%

phrases related to political statements or actions

commas and phrases indicating shifts in the narrative or context

No Comments

No Known Activations