INDEX

Explanations

terms related to human emotional or behavioral responses, specifically focusing on reactions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "reaction."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

locked

-0.73

holder

-0.73

hold

-0.73

 Sinai

-0.70

iciency

-0.66

 Ethiopian

-0.65

enture

-0.64

red

-0.63

Hidden

-0.61

 Sustainable

-0.61

POSITIVE LOGITS

ivated

1.24

ivation

1.21

 reactions

1.18

 reaction

1.15

ivating

1.15

aries

1.08

 Reaction

1.01

naires

0.86

ively

0.84

iv

0.83

Activations Density 0.029%

terms related to human emotional or behavioral responses, specifically focusing on reactions

instances of the word "reaction."

No Comments

No Known Activations