INDEX

Explanations

references to hills or elevated terrains

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

toMatchSnapshot

-0.72

 appa

-0.69

quard

-0.68

MFC

-0.67

ronesia

-0.67

),

-0.67

Kaieteur

-0.66

sko

-0.65

})}

-0.64

retan

-0.64

POSITIVE LOGITS

 Hill

2.35

 hill

2.26

Hill

2.17

 HILL

2.15

 hills

1.94

hill

1.90

 Hills

1.71

HILL

1.69

Hills

1.52

 HILLS

1.46

Activations Density 0.040%

references to hills or elevated terrains

No Comments

No Known Activations