INDEX

Explanations

connections between concepts, with an emphasis on specific terms related to condition, targeting, and action

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

Personendaten

-0.64

ArgumentParser

-0.52

interopRequire

-0.51

msgTypes

-0.50

FunctionFlags

-0.49

 aDecoder

-0.48

ViewInit

-0.45

MessageState

-0.44

 Normdatei

-0.44

ServletException

-0.44

POSITIVE LOGITS

 émissions

0.43

webElementXpaths

0.42

 cérami

0.41

 mijne

0.40

 recomienda

0.40

 zijne

0.39

 céramique

0.39

 útiles

0.39

 inwest

0.39

 cremallera

0.39

Activations Density 0.596%

connections between concepts, with an emphasis on specific terms related to condition, targeting, and action

No Comments

No Known Activations