INDEX

Explanations

phrases that indicate definitions or explanations

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

 Eder

-0.46

 없습니다

-0.41

 junger

-0.41

 jonge

-0.40

没有任何

-0.39

 liste

-0.38

 Former

-0.38

Jody

-0.37

 Fiske

-0.37

 students

-0.36

POSITIVE LOGITS

 Means

1.09

Means

1.05

means

1.00

 MEANS

0.95

 mean

0.95

 Mean

0.95

 MEAN

0.94

Mean

0.92

mean

0.90

 means

0.89

Activations Density 0.105%

phrases that indicate definitions or explanations

No Comments

No Known Activations