INDEX

Explanations

words indicating frequency or exceptions to rules

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Roskov

-0.67

GraphicsUnit

-0.66

անո

-0.63

SuppressMessage

-0.61

 onData

-0.60

 تضيفلها

-0.59

withIdentifier

-0.59

حوالہ

-0.59

ślę

-0.59

CodeAttribute

-0.58

POSITIVE LOGITS

 being

0.87

being

0.66

 BEING

0.60

 censiti

0.57

 having

0.56

 fiind

0.53

 Being

0.52

Being

0.50

رشف

0.50

iliki

0.49

Activations Density 0.242%

words indicating frequency or exceptions to rules

No Comments

No Known Activations