INDEX

Explanations

references to the breadth or diversity of topics or subjects

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

 Zimmermann

-0.71

 pulumi

-0.68

CellValue

-0.68

htbp

-0.67

 trajets

-0.66

 Metallurgy

-0.66

èdia

-0.66

rentino

-0.65

CastException

-0.64

ashvili

-0.63

POSITIVE LOGITS

 wide

2.70

 Wide

2.54

Wide

2.38

 WIDE

2.33

wide

2.14

WIDE

2.11

 widest

2.02

 broad

1.71

 wider

1.71

wider

1.55

Activations Density 0.078%

references to the breadth or diversity of topics or subjects

No Comments

No Known Activations