INDEX

Explanations

instances of the word "em" followed by a number, possibly indicating a reference to a measurement or value

oai_token-act-pair · gpt-3.5-turbo

instances of the term "em" indicating emotional expressions or sentiments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 otherwise

-0.65

bed

-0.60

 enriched

-0.60

 stranger

-0.58

SourceFile

-0.58

 yours

-0.57

 congratulations

-0.56

 Kubrick

-0.56

 retard

-0.56

 redistributed

-0.56

POSITIVE LOGITS

peror

1.21

issions

1.21

achine

1.15

poral

1.10

icro

1.08

psey

1.07

bley

1.07

otional

1.07

phis

1.07

blance

1.06

Activations Density 0.033%

instances of the word **"em"** followed by a number, possibly indicating a reference to a measurement or value

instances of the term "em" indicating emotional expressions or sentiments

No Comments

No Known Activations

instances of the word "em" followed by a number, possibly indicating a reference to a measurement or value