INDEX

Explanations

phrases suggesting speculation or uncertainty

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

wait

-0.69

anding

-0.69

osi

-0.67

catentry

-0.65

aris

-0.62

etting

-0.59

 plaintiff

-0.59

center

-0.59

ilant

-0.58

leaf

-0.58

POSITIVE LOGITS

 been

1.39

 gotten

1.33

been

1.20

gotten

1.12

 undergone

1.10

 gone

1.10

 Been

1.07

 originated

1.06

 begun

1.04

 contributed

1.03

Activations Density 0.091%

phrases suggesting speculation or uncertainty

No Comments

No Known Activations