INDEX

Explanations

verbs that indicate watching or choosing something, as well as general connecting words

oai_token-act-pair · gemini-2.0-flash

watch

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_25/width_16k/average_l0_41

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.25.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

 watch

-2.47

 Watch

-2.20

watch

-2.16

Watch

-2.09

 watching

-2.02

 watches

-2.02

 WATCH

-2.00

 watched

-1.98

 Watching

-1.84

watching

-1.81

POSITIVE LOGITS

новниш

0.60

complexContent

0.56

jooq

0.52

WithIOException

0.50

Personensuche

0.49

 opérateur

0.49

featureID

0.49

 évaluations

0.48

autoconfigure

0.48

 Himo

0.48

Activations Density 4.578%

verbs that indicate watching or choosing something, as well as general connecting words

watch

No Comments

No Known Activations