INDEX

Explanations

references to legal or political authority and processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_11/width_16k/average_l0_80

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

 lắm

-0.51

loss

-0.47

 circunstan

-0.46

 comples

-0.45

 loosing

-0.45

closePath

-0.44

eled

-0.43

เสีย

-0.43

Poloha

-0.43

entamiento

-0.42

POSITIVE LOGITS

 initComponents

0.77

 disobedience

0.73

 protest

0.71

Javadoc

0.69

protest

0.67

 protested

0.66

 dared

0.65

 disobey

0.64

 protests

0.63

Activations Density 0.362%

references to legal or political authority and processes

No Comments

No Known Activations