INDEX

Explanations

tokens related to numerical or coded information

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

numbers followed by }

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

}]);

-0.35

帖最后由

-0.35

 Capacidad

-0.35

 acepción

-0.35

 colgantes

-0.34

 felizes

-0.34

 Potencia

-0.34

 aldri

-0.33

 identidade

-0.33

 navegando

-0.33

POSITIVE LOGITS

MigrationBuilder

0.75

CloseOperation

0.66

WithIOException

0.66

DockStyle

0.61

انيف

0.57

 المعيارى

0.56

richTextPanel

0.55

tonode

0.54

 snippetHide

0.54

 صوتيه

0.52

Activations Density 0.133%

tokens related to numerical or coded information

numbers followed by }

No Comments

No Known Activations