INDEX

Explanations

constructs related to computational tasks and processes

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

RegressionTest

-0.65

 &___

-0.65

-0.63

 CanadaChoose

-0.51

 betweenstory

-0.50

 հղումներ

-0.49

 invokingState

-0.49

GEBURTSDATUM

-0.49

">//

-0.48

цездатний

-0.48

POSITIVE LOGITS

openConnection

0.41

OGND

0.40

enosine

0.40

 للاسماء

0.40

 وتسجيلات

0.39

FunctionFlags

0.39

méno

0.39

 dég

0.39

 molecule

0.36

petto

0.36

Activations Density 0.895%

constructs related to computational tasks and processes

No Comments

No Known Activations