INDEX

Explanations

references to focus and emphasis in discussions about processes or systems

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

 caminhada

-0.42

 conclusão

-0.40

 føl

-0.40

úsculo

-0.40

voerd

-0.37

 Moscú

-0.36

 llegue

-0.35

 garrafa

-0.35

 talvez

-0.35

eseorang

-0.35

POSITIVE LOGITS

AnchorStyles

0.66

AxisAlignment

0.65

 autorytatywna

0.64

 Emphasis

0.62

 emphasis

0.62

 emphasizing

0.62

 emphasize

0.59

重視

0.58

ьаж

0.58

OGND

0.57

Activations Density 0.016%

references to focus and emphasis in discussions about processes or systems

No Comments

No Known Activations