INDEX

Explanations

references to 'subjects' or 'topics' in academic or formal contexts

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

:✨

-0.71

GIVEREF

-0.69

úgó

-0.66

 queſta

-0.65

ейчас

-0.64

ThroughAttribute

-0.63

󠁮

-0.63

niſſe

-0.63

 パンチラ

-0.62

 sanitaire

-0.62

POSITIVE LOGITS

 subject

1.03

 topic

1.02

 Subject

0.96

subject

0.92

 theme

0.92

Subject

0.91

 Topic

0.91

Topic

0.91

topic

0.88

 Theme

0.82

Activations Density 0.092%

references to 'subjects' or 'topics' in academic or formal contexts

No Comments

No Known Activations