INDEX

Explanations

phrases indicating contradictions or comparisons

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

 myſelf

-0.64

Bakgrunnsstoff

-0.56

 ostavi

-0.52

 ComVisible

-0.52

Ecotoxicity

-0.52

 textStatus

-0.52

queryInterface

-0.48

 resor

-0.48

 Privilege

-0.47

WebElementEntity

-0.47

POSITIVE LOGITS

 nawet

0.40

 defStyle

0.38

 even

0.38

/**

0.38

 даже

0.37

SourceChecksum

0.37

 bahkan

0.36

ecap

0.35

 gezet

0.35

 defaultstate

0.34

Activations Density 0.069%

phrases indicating contradictions or comparisons

No Comments

No Known Activations