INDEX

Explanations

terms related to emergencies and safety assessments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 architectures

-0.34

 history

-0.33

 fame

-0.33

 shame

-0.32

History

-0.32

ByExample

-0.31

直

-0.31

町

-0.31

下

-0.30

 camicia

-0.30

POSITIVE LOGITS

 poffe

0.63

 الحره

0.63

 محفوظة

0.62

 AssemblyCulture

0.61

ſelves

0.60

 AssemblyVersion

0.60

 prelimin

0.59

 ब्रेकडाउन

0.54

ſelf

0.53

 للاسماء

0.52

Activations Density 0.064%

terms related to emergencies and safety assessments

No Comments

No Known Activations