INDEX

Explanations

identifiers and names used in code

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

elemField

-0.58

findpost

-0.56

ftagPool

-0.47

 AppCompatTheme

-0.45

 approximately

-0.43

 Approximately

-0.39

 defense

-0.37

 saved

-0.37

nologue

-0.37

 recovered

-0.36

POSITIVE LOGITS

 disambiguazione

0.73

WebServlet

0.56

 ujednoznacz

0.55

tagHelperRunner

0.50

 kasarigan

0.49

 identidade

0.48

NewGuid

0.48

 ramai

0.47

getSimpleName

0.47

 orejas

0.46

Activations Density 0.021%

identifiers and names used in code

No Comments

No Known Activations