INDEX

Explanations

references to manual control or manual operations in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_1/width_16k/average_l0_69

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.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

 autorytatywna

-0.50

acamata

-0.49

 sentimenti

-0.46

IBLIO

-0.43

Elsewhere

-0.43

utilisons

-0.43

 telefónica

-0.42

 MenuView

-0.41

 berdua

-0.41

ểm

-0.40

POSITIVE LOGITS

 manual

0.78

 capture

0.70

 Capture

0.66

Capture

0.64

manual

0.64

capture

0.59

 manuals

0.59

 Manual

0.59

Manual

0.59

 manually

0.57

Activations Density 0.129%

references to manual control or manual operations in various contexts

No Comments

No Known Activations