INDEX

Explanations

proper names and individuals mentioned in the text

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

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

 సౌకర్య

-0.39

 Verhältnisse

-0.38

配

-0.37

Rhestr

-0.35

SceneManagement

-0.34

 Потому

-0.34

 μο

-0.34

 poroz

-0.33

 кӀ

-0.33

 remarkably

-0.33

POSITIVE LOGITS

enderror

0.55

iſten

0.54

 ainfi

0.48

ſcher

0.48

 plufieurs

0.47

$_"

0.46

 vicenda

0.46

 الحره

0.45

étoit

0.44

脚注の使い方

0.44

Activations Density 0.016%

proper names and individuals mentioned in the text

No Comments

No Known Activations