INDEX

Explanations

references to political figures and organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

baseball

-0.41

 obro

-0.39

 berpikir

-0.38

PyExc

-0.36

tagHelperRunner

-0.36

 django

-0.35

 Tbilisi

-0.32

sø

-0.32

StartTag

-0.32

DeleteBehavior

-0.32

POSITIVE LOGITS

 Lumpur

0.77

 Malaysian

0.76

 Malaysia

0.75

Malaysia

0.74

UALA

0.63

Malay

0.63

AYSIA

0.60

 malaysia

0.58

BASEPATH

0.58

 Johor

0.57

Activations Density 0.256%

references to political figures and organizations

No Comments

No Known Activations