INDEX

Explanations

references to vocal performance and vocal-related terminology

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Bush

-0.56

 money

-0.54

estimate

-0.52

estimation

-0.50

slant

-0.49

 Rams

-0.49

 Ramsey

-0.48

 isomorphism

-0.48

 Sands

-0.48

 Kits

-0.48

POSITIVE LOGITS

Vocal

1.02

vocal

0.93

 Vocal

0.92

 vocal

0.86

 Vocals

0.75

 vokal

0.75

 vocals

0.73

 vocales

0.71

Voc

0.67

 vocalist

0.66

Activations Density 0.003%

references to vocal performance and vocal-related terminology

No Comments

No Known Activations