INDEX

Explanations

the term "specific" as it relates to evaluation or measurement

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

specific

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

Negative Logits

 hablando

-0.56

days

-0.47

eadas

-0.46

ర్య

-0.46

 polyps

-0.44

razor

-0.43

jelaskan

-0.42

OrEqual

-0.42

 márm

-0.42

 giudi

-0.42

POSITIVE LOGITS

 Normdatei

0.98

RectangleBorder

0.80

ItemBackground

0.76

SpringBootTest

0.75

PerformLayout

0.74

>");

0.71

BURGH

0.71

 gravity

0.71

UnsafeEnabled

0.70

elemField

0.68

Activations Density 0.065%

the term "specific" as it relates to evaluation or measurement

specific

No Known Activations