INDEX

Explanations

conversational phrases that indicate agreement or acknowledgment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

 Roskov

-0.96

 saites

-0.78

 BoxFit

-0.77

 Cæsar

-0.73

Cordialement

-0.73

abyrinth

-0.72

 marquis

-0.72

genheim

-0.72

endpush

-0.72

epa

-0.71

POSITIVE LOGITS

 Well

1.05

Well

0.99

 well

0.74

hey

0.71

 WELL

0.68

well

0.66

 technically

0.65

 Wells

0.65

 уж

0.64

Pues

0.64

Activations Density 0.032%

conversational phrases that indicate agreement or acknowledgment

No Comments

No Known Activations