INDEX

Explanations

numerical values and statistical terms that indicate comparisons and results in studies

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

 snippetHide

-0.55

 Taktlose

-0.55

imer

-0.49

GTCX

-0.49

 PeEnEo

-0.49

SequentialGroup

-0.48

XmlAccessorType

-0.48

Lol

-0.47

 Conven

-0.47

 ***!

-0.46

POSITIVE LOGITS

ďaka

0.47

 attributed

0.44

 terciopelo

0.41

 debido

0.40

 devido

0.39

 Hinweis

0.38

 głowy

0.38

 rahasia

0.38

 Erscheinung

0.38

 Zufall

0.37

Activations Density 0.946%

numerical values and statistical terms that indicate comparisons and results in studies

No Comments

No Known Activations