INDEX

Explanations

references to specific subjects or items that are prevalent or widely available

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 33-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.33.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

ativement

-0.41

intios

-0.40

[]:

-0.35

parsedMessage

-0.35

featureID

-0.34

delwed

-0.34

 мастер

-0.34

 BoxFit

-0.33

nodoc

-0.33

yntaxException

-0.33

POSITIVE LOGITS

out

1.50

 draußen

0.93

Out

0.79

Out

0.78

out

0.77

 afuera

0.67

ออกมา

0.66

OUT

0.66

 floating

0.65

 fuori

0.62

Activations Density 0.250%

references to specific subjects or items that are prevalent or widely available

No Comments

No Known Activations