INDEX

Explanations

dates and locations related to political events

oai_token-act-pair · gpt-3.5-turbo

occurrences of slashes or dividers, likely indicating sections or elements in a list or formatting in the text

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

kson

-0.73

 guid

-0.66

mie

-0.64

UDP

-0.63

racuse

-0.62

IPM

-0.61

 converted

-0.60

 Baldwin

-0.60

 Mama

-0.58

 ranges

-0.56

POSITIVE LOGITS

¡

0.76

Art

0.73

HCR

0.71

Gallery

0.70

Appearances

0.69

âĹ¼

0.66

Film

0.65

OUR

0.65

¥

0.65

¶

0.64

Activations Density 0.022%

dates and locations related to political events

occurrences of slashes or dividers, likely indicating sections or elements in a list or formatting in the text

No Comments

No Known Activations