INDEX

Explanations

phrases indicating causation or implication

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

 HttpNotFound

-0.56

 ThemeData

-0.55

windowFixed

-0.52

 plunger

-0.50

featureID

-0.49

tagext

-0.48

SourceChecksum

-0.47

 transférez

-0.47

atro

-0.47

WireFormatLite

-0.47

POSITIVE LOGITS

 pelúcia

0.43

それだけ

0.39

 initComponents

0.36

0.35

[]:

0.35

 commandement

0.35

 становника

0.34

 acrylique

0.34

 conclusión

0.34

 fisuras

0.34

Activations Density 0.691%

phrases indicating causation or implication

No Comments

No Known Activations