INDEX

Explanations

phrases indicating personal experiences and interactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hur

-0.39

tom

-0.35

 ContentValues

-0.34

 provision

-0.33

即

-0.33

宣

-0.33

✅

-0.32

createElement

-0.32

Hur

-0.32

洲

-0.31

POSITIVE LOGITS

 Signalez

0.67

 hieronder

0.66

AndroidJUnit

0.60

majánló

0.56

 betweenstory

0.56

 morire

0.54

 AssemblyCulture

0.53

 miniaturka

0.53

Below

0.51

 męska

0.51

Activations Density 0.083%

phrases indicating personal experiences and interactions

No Comments

No Known Activations