INDEX

Explanations

phrases related to opinions or perspectives

oai_token-act-pair · gpt-3.5-turbo

phrases related to public sentiment and policy discussion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Flavoring

-0.71

éŃĶ

-0.66

kefeller

-0.64

ickets

-0.60

Hau

-0.59

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.57

onductor

-0.57

 fibre

-0.56

racuse

-0.56

Merit

-0.56

POSITIVE LOGITS

!?

0.81

????????

0.81

????

0.80

?!

0.76

???

0.76

âĶĢâĶĢâĶĢâĶĢ

0.75

huh

0.73

...?

0.73

plet

0.71

"?

0.71

Activations Density 1.122%

phrases related to opinions or perspectives

phrases related to public sentiment and policy discussion

No Comments

No Known Activations