INDEX

Explanations

code-related tokens, especially method and format names with "toLowercase"

oai_token-act-pair · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

parsedMessage

-0.96

oredCriteria

-0.93

haustible

-0.84

}{*}{}

-0.79

æa

-0.77

unknownFields

-0.75

encils

-0.75

æus

-0.75

ernalia

-0.74

ebvre

-0.74

POSITIVE LOGITS

<bos>

0.47

enumi

0.45

 térm

0.40

 florales

0.40

 extré

0.38

 Ejecutivo

0.35

 éta

0.35

 précie

0.34

 initComponents

0.34

 éché

0.34

Activations Density 9.490%

code-related tokens, especially method and format names with "toLowercase"

No Comments

No Known Activations