INDEX

Explanations

physical objects and actions involving them

oai_token-act-pair · gpt-3.5-turbo

items and references related to weapons and firearms

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

lees

-0.86

ablishment

-0.85

agonists

-0.82

attacks

-0.81

WAYS

-0.79

Init

-0.78

 collisions

-0.78

forces

-0.75

INESS

-0.75

 behaviours

-0.75

POSITIVE LOGITS

 suitcase

1.33

 notebook

1.32

 bottle

1.28

 clipboard

1.23

 cassette

1.21

 booklet

1.19

 cardboard

1.18

 pamphlet

1.17

jar

1.14

 laptop

1.14

Activations Density 0.426%

physical objects and actions involving them

items and references related to weapons and firearms

No Comments

No Known Activations