INDEX

Explanations

phrases related to health and medical advice

oai_token-act-pair · gpt-3.5-turbo

phrases related to significant experiences and emotional impact

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 Malays

-0.71

Scroll

-0.68

 Spice

-0.65

 Patt

-0.64

 Malaysia

-0.63

 Bian

-0.62

 McKenna

-0.61

 paperback

-0.61

Mia

-0.61

 Hogwarts

-0.60

POSITIVE LOGITS

 incent

0.81

amount

0.79

entimes

0.77

 behavi

0.75

 respectively

0.74

SourceFile

0.71

 tremend

0.69

aldi

0.69

 streng

0.68

ç¥ŀ

0.66

Activations Density 0.579%

phrases related to health and medical advice

phrases related to significant experiences and emotional impact

No Comments

No Known Activations