INDEX

Explanations

references to visual perception or observation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 loisirs

-0.79

 Hofmann

-0.78

 Wadsworth

-0.75

 universitarios

-0.74

 Hoffman

-0.73

knapp

-0.73

Kort

-0.72

shaft

-0.71

Grath

-0.70

vedev

-0.69

POSITIVE LOGITS

see

2.13

 sees

2.04

See

1.95

SEE

1.94

see

1.88

 Sees

1.88

See

1.87

 seeing

1.73

SEE

1.69

saw

1.67

Activations Density 0.097%

references to visual perception or observation

No Comments

No Known Activations

references to visual perception or observation

No Comments

No Known Activations