INDEX

Explanations

phrases related to boosting or increasing something, such as performance, numbers, or prices

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

mire

-1.02

ocument

-0.75

ollow

-0.74

wic

-0.74

irez

-0.72

":[

-0.71

cloth

-0.71

olina

-0.71

lace

-0.70

 Ended

-0.70

POSITIVE LOGITS

own

1.42

 chances

1.36

 ability

1.26

 credibility

1.22

 reputation

1.13

 stature

1.12

 effectiveness

1.10

 competitiveness

1.08

 productivity

1.08

 attractiveness

1.07

Activations Density 0.191%

phrases related to boosting or increasing something, such as performance, numbers, or prices

No Comments

No Known Activations