INDEX

Explanations

references to layers and structures in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 للاسماء

-0.60

/**

-0.47

tagHelperRunner

-0.44

VIAF

-0.42

----</

-0.41

 كومونز

-0.41

Eater

-0.39

 Roskov

-0.39

curs

-0.39

 Stable

-0.39

POSITIVE LOGITS

 berdua

0.51

 beiden

0.47

 astore

0.42

 obrazy

0.42

\}_{

0.42

ViewFeatures

0.41

 båda

0.40

 Baillargeon

0.39

 begge

0.39

 opinião

0.39

Activations Density 1.731%

references to layers and structures in various contexts

No Comments

No Known Activations