INDEX

Explanations

phrases indicating a significant emphasis or importance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 11-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_11/width_16k/average_l0_118

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.11.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

 <<<<<<<<<<<<<<

-0.93

FieldBuilder

-0.90

principalColumn

-0.88



-0.88

SharedDtor

-0.87

<unused43>

-0.82

 للاسماء

-0.82

<unused42>

-0.81

<unused41>

-0.81

<pad>

-0.81

POSITIVE LOGITS

the

0.65

The

0.41

The

0.41

 those

0.41

那些

0.38

jalá

0.38

 there

0.37

Those

0.36

Sometimes

0.35

Some

0.34

Activations Density 0.381%

phrases indicating a significant emphasis or importance

No Comments

No Known Activations