INDEX

Explanations

references to specific locations or places

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

ignoire

-0.50

 onStart

-0.50

 Navarra

-0.49

Kim

-0.48

 hydrodynamic

-0.48

 biometric

-0.47

 Hydro

-0.47

Hy

-0.47

 Monmouth

-0.47

 Harwood

-0.46

POSITIVE LOGITS

place

1.15

 PLACE

1.11

PLACE

1.08

Place

1.06

 Place

1.05

 place

1.02

places

0.97

Places

0.95

 Places

0.91

 places

0.89

Activations Density 0.056%

references to specific locations or places

No Comments

No Known Activations