INDEX

Explanations

Words related to surprising or shocking events

oai_token-act-pair · gpt-3.5-turbo

references to emotional responses characterized as "shock."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

Ago

-0.79

Community

-0.77

yards

-0.75

abouts

-0.74

aligned

-0.71

yard

-0.68

Ver

-0.65

Cod

-0.65

adra

-0.65

aler

-0.64

POSITIVE LOGITS

 shock

4.03

 shocks

2.60

shock

2.54

 Shock

2.28

 surprise

2.19

Shock

1.92

 shocked

1.74

 aston

1.52

 disbelief

1.42

 stunned

1.38

Activations Density 0.019%

Words related to surprising or shocking events

references to emotional responses characterized as "shock."

No Comments

No Known Activations