INDEX

Explanations

boolean conjunctions and indications of conditions within 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 @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

Autoritní

-0.47

 Paglinawan

-0.46

adaptiveStyles

-0.43

әрмәләр

-0.42

Rujuakan

-0.41

 desmotivaciones

-0.40

SharedCtor

-0.39

HideFlags

-0.39

 мәкал

-0.38

 générations

-0.37

POSITIVE LOGITS

 also

0.62

de

0.60

auri

0.59

noh

0.59

 dera

0.58

ARO

0.57

is

0.57

 için

0.54

ERD

0.54

MSA

0.54

Activations Density 1.786%

boolean conjunctions and indications of conditions within various contexts

No Comments

No Known Activations