INDEX

Explanations

phrases indicating extraction or uncovering of information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

 "..\..\

-0.47

 reputación

-0.45

 Grüsse

-0.45

epam

-0.41

 Beobachtung

-0.40

 alrededores

-0.40

 autorytatywna

-0.40

 autorité

-0.40

 afstand

-0.39

andangan

-0.39

POSITIVE LOGITS

 mobilization

0.59

mobili

0.56

掏

0.51

 invocation

0.50

 mobili

0.49

dug

0.49

 mobilized

0.49

BufferException

0.48

 mobilize

0.47

拿出

0.47

Activations Density 0.792%

phrases indicating extraction or uncovering of information

No Comments

No Known Activations