INDEX

Explanations

occurrences of the word "uses."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_mlp_out

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

 uses

-2.06

 Uses

-1.93

Uses

-1.80

uses

-1.46

 USES

-1.41

 utilizes

-1.02

 usos

-0.94

USES

-0.93

 utilizza

-0.86

 använder

-0.83

POSITIVE LOGITS

 InputDecoration

0.77

 chi̍t

0.75

RectangleBorder

0.72

+#+#

0.68

 uncomment

0.63

 Roskov

0.57

 disambiguazione

0.56

 ISNI

0.56

Personendaten

0.56

 Wicidata

0.55

Activations Density 0.089%

occurrences of the word "uses."

No Comments

No Known Activations