INDEX

Explanations

details related to objects held in hands

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

"hand" or related words

np_acts-logits-general · gemini-2.0-flash

hand or hands after 'in'

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

NewRow

-0.43

ppelin

-0.42

 entry

-0.41

Entry

-0.41

bitat

-0.40

 Betten

-0.40

 Chairs

-0.39

cd

-0.38

entry

-0.38

AGON

-0.38

POSITIVE LOGITS

 hand

1.40

 Hand

1.30

hand

1.28

Hand

1.24

 hands

1.23

 HAND

1.23

HAND

1.19

 handed

1.18

hands

1.17

 handheld

1.15

Activations Density 0.179%

details related to objects held in hands

"hand" or related words

hand or hands after 'in'

No Comments

No Known Activations