INDEX

Explanations

phrases related to interactions or opinions involving other people

oai_token-act-pair · gpt-3.5-turbo

references to social interactions and perceptions about other individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

Accessory

-0.90

ouf

-0.65

 Humane

-0.63

Arsenal

-0.62

 Utility

-0.59

wagen

-0.59

 Luck

-0.58

LOCK

-0.58

HUN

-0.57

 Heist

-0.56

POSITIVE LOGITS

worldly

0.97

 besides

0.93

 describ

0.82

iations

0.74

than

0.73

 than

0.70

paces

0.69

 perspectives

0.69

 equally

0.69

 avenues

0.65

Activations Density 0.177%

phrases related to interactions or opinions involving other people

references to social interactions and perceptions about other individuals

No Comments

No Known Activations