INDEX

Explanations

phrases related to a specific point of view or perspective

oai_token-act-pair · gpt-3.5-turbo

phrases that express different perspectives or points of view

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

cffff

-0.70

 corrid

-0.66

kefeller

-0.63

wen

-0.62

agh

-0.61

pop

-0.60

aucuses

-0.60

 tiss

-0.60

 seaf

-0.60

enge

-0.59

POSITIVE LOGITS

finder

0.87

aries

0.85

atics

0.80

ership

0.71

ibility

0.70

 angles

0.68

 perspectives

0.68

models

0.68

 mirrors

0.67

oji

0.66

Activations Density 0.014%

phrases related to a specific point of view or perspective

phrases that express different perspectives or points of view

No Comments

No Known Activations