INDEX

Explanations

terms related to statistical analysis and data classification

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

 nakalista

-0.43

 Verſ

-0.38

AccessorTable

-0.38

 quaisquer

-0.37

 CURIAM

-0.36

 faſt

-0.36

 cauſe

-0.35

 houſe

-0.35

 Anſ

-0.34

 feroit

-0.33

POSITIVE LOGITS

AddTagHelper

0.63

versario

0.53

 Lingkungan

0.52

Shallow

0.51

ículos

0.50

boutin

0.48

 Alten

0.47

Solución

0.47

immel

0.47

 Erden

0.47

Activations Density 0.414%

terms related to statistical analysis and data classification

No Comments

No Known Activations