INDEX

Explanations

discussions of proposed methods and algorithms, particularly in the context of their numerical properties and implementations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

AnchorStyles

-0.66

ocity

-0.63

 répé

-0.63

XtraEditors

-0.61

 noDo

-0.60

kì

-0.60

dox

-0.59



-0.59

risse

-0.59

—

-0.58

POSITIVE LOGITS

thâu

0.54

 tackles

0.53

 designed

0.53

 Belgien

0.52

 developed

0.52

 innovative

0.50

ujudkan

0.50

Aktualisiert

0.49

GORITHM

0.49

Viitteet

0.47

Activations Density 0.142%

discussions of proposed methods and algorithms, particularly in the context of their numerical properties and implementations

No Comments

No Known Activations