INDEX

Explanations

references to rows in a structured data context, such as a spreadsheet or database

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

 majeur

-0.88

 Monfieur

-0.87

 Theſe

-0.81

DispatchToProps

-0.77

 poffible

-0.77

 virtù

-0.76

 enfans

-0.76

 Majefty

-0.75

 doskon

-0.75

 fatis

-0.75

POSITIVE LOGITS

row

2.71

Row

2.70

Row

2.49

 rows

2.39

row

2.29

ROW

2.20

 Rows

2.13

ROW

2.05

rows

2.00

Rows

1.97

Activations Density 0.028%

references to rows in a structured data context, such as a spreadsheet or database

No Comments

No Known Activations