INDEX

Explanations

instances of the word "filling" and its variations, indicating actions related to filling something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_65k/average_l0_73

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.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

SupportActionBar

-0.64

########.

-0.62

 religieuses

-0.60

parsedMessage

-0.60

providedIn

-0.60

 SwitchCompat

-0.60

 démocr

-0.59

 chromat

-0.59

 invokingState

-0.59

tagHelperRunner

-0.58

POSITIVE LOGITS

 fill

2.52

 filling

2.45

 Fill

2.31

 FILL

2.27

Fill

2.23

 Filling

2.23

Filling

2.20

fill

2.20

 filled

2.17

 empty

2.16

Activations Density 0.043%

instances of the word "filling" and its variations, indicating actions related to filling something

No Comments

No Known Activations