INDEX

Explanations

phrases indicating the presence or implementation of established procedures or systems

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

gee

-0.76

etti

-0.72

anche

-0.71

yssey

-0.70

odge

-0.67

 DRAG

-0.66

zzy

-0.66

ourge

-0.65

 Ging

-0.65

zees

-0.64

POSITIVE LOGITS

 whereby

0.83

bos

0.80

alities

0.72

ifice

0.64

atives

0.64

 incentiv

0.64

 antioxid

0.61

 protections

0.61

bodied

0.60

ative

0.59

Activations Density 0.031%

phrases indicating the presence or implementation of established procedures or systems

No Comments

No Known Activations