INDEX

Explanations

the word "traditional", but sometimes other words nearby get a low activation value

oai_token-act-pair · gemini-2.0-flash

traditional

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_0/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.0.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

 real

-0.84

'),

-0.82

 high

-0.77

 very

-0.76

non

-0.75

one

-0.74

"),

-0.74

')

-0.73

'],

-0.73

 traditional

-0.73

POSITIVE LOGITS

 normaux

1.35

 définiti

1.30

 démocr

1.30

 fermés

1.25

 feroit

1.24

 déchir

1.22

 supérieurs

1.18

 auroit

1.17

 pouvoit

1.16

 mauva

1.16

Activations Density 3.107%

the word "traditional", but sometimes other words nearby get a low activation value

traditional

No Comments

No Known Activations