INDEX

Explanations

proper nouns related to locations, titles, or specific entities

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

 almofada

-0.44

AnchorStyles

-0.43

AddHtmlAttribute

-0.39

@"/

-0.39

 godine

-0.39

ratulations

-0.38

AxisAlignment

-0.38

Portale

-0.38

 województ

-0.37

 França

-0.37

POSITIVE LOGITS

tagHelperRunner

0.59

contentLoaded

0.53

isContained

0.52

⟬

0.50

 ✭✭

0.50

ьажоргаш

0.49

 հղումներ

0.48

:✨

0.47

 endforeach

0.46

saraba

0.45

Activations Density 0.084%

proper nouns related to locations, titles, or specific entities

No Comments

No Known Activations