INDEX

Explanations

words that describe body parts, particularly arms, legs, or fingers, often in a disturbing way

oai_token-act-pair · gemini-2.0-flash

variations of the word "cross"

np_token-act-pair-logits · gpt-4o-mini

cross

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Anſ

-0.71

RTEX

-0.69

Efq

-0.63

 Inſ

-0.62

 Italijanski

-0.60

 ſche

-0.60

 صوتيه

-0.59

endphp

-0.59

DeleteBehavior

-0.59

 himſelf

-0.58

POSITIVE LOGITS

 crossed

1.24

 cross

1.16

cross

1.04

 CROSS

1.04

Cross

1.02

crossed

1.00

 Cross

1.00

 crossing

0.93

 crosses

0.93

crossing

0.83

Activations Density 3.151%

words that describe body parts, particularly arms, legs, or fingers, often in a disturbing way

variations of the word "cross"

cross

No Comments

No Known Activations