INDEX

Explanations

expressions of strong positive opinions or emotions

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

ending punctuation or determiners

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

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

OGND

-0.84

 محفوظة

-0.79

yntaxException

-0.71

ſelf

-0.70

 Administrativna

-0.68

 cherchés

-0.68

Personensuche

-0.67

URLException

-0.67

ंदीखरीदारी

-0.66

 يتيمه

-0.65

POSITIVE LOGITS

↵

0.40

↵↵

0.35

0.34

 Warbler

0.34

 blew

0.34

 rounding

0.31

 nice

0.31

0.29

 Rund

0.29

Activations Density 0.002%

expressions of strong positive opinions or emotions

ending punctuation or determiners

No Comments

No Known Activations