INDEX

Explanations

mentions of specific names and programming-related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.12.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

Rha

-0.68

 Utica

-0.67

 NTIS

-0.66

 withRouter

-0.65

 Teri

-0.65

 Bork

-0.62

 Marietta

-0.61

 Kine

-0.61

Mox

-0.61

 Braxton

-0.60

POSITIVE LOGITS

Jag

0.96

 Duncan

0.95

Duncan

0.93

 Campbell

0.93

 Walkover

0.91

 Jaguar

0.91

Jaguar

0.91

hasMoreElements

0.88

олові

0.88

Datuak

0.87

Activations Density 0.035%

mentions of specific names and programming-related terms

No Comments

No Known Activations