INDEX

Explanations

phrases related to accusations or suspicions involving named individuals

oai_token-act-pair · gpt-3.5-turbo

references to Donald Trump

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

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

ruary

-0.81

dos

-0.69

ãĥ¼ãĥ«

-0.66

dden

-0.65

glers

-0.64

ariat

-0.63

ibility

-0.63

Thumbnail

-0.63

gers

-0.62

zsche

-0.61

POSITIVE LOGITS

 Recomm

0.73

Senate

0.61

 diabetic

0.59

 allergic

0.58

 optimistic

0.56

Ford

0.56

 palace

0.55

 overrun

0.55

 holiest

0.55

Intel

0.54

Activations Density 0.033%

phrases related to accusations or suspicions involving named individuals

references to Donald Trump

No Comments

No Known Activations