INDEX

Explanations

instances of negation and expressions of preference

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Autoritní

-0.66

:✨

-0.58

ItemBackground

-0.57

yarnpkg

-0.57

########.

-0.54

 PyLong

-0.52

 zám

-0.52

 disambiguazione

-0.52

 nahilalakip

-0.50

DockStyle

-0.47

POSITIVE LOGITS

évaluateur

0.39

 automatically

0.37

 Expression

0.36

 चीज़ों

0.35

 expressions

0.35

 language

0.34

 expression

0.34

jah

0.34

 Expressions

0.33

 vocabulary

0.32

Activations Density 0.028%

instances of negation and expressions of preference

No Comments

No Known Activations