INDEX

Explanations

terms related to statistical models and analysis

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

følgelig

-0.47

 Audiodateien

-0.45

 hånd

-0.44

ſelves

-0.44

 الحره

-0.41

kjø

-0.41

 jScrollPane

-0.40

iastes

-0.39

 skolen

-0.39

 daugiau

-0.38

POSITIVE LOGITS

 variant

0.98

 Variant

0.90

variant

0.89

styleType

0.79

 variants

0.74

 BoxDecoration

0.73

 Variants

0.73

Variant

0.72

Variants

0.69

 variation

0.67

Activations Density 0.004%

terms related to statistical models and analysis

No Comments

No Known Activations