INDEX

Explanations

references to deception, manipulation, and falsehoods

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_0/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.0.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

 zurückge

-0.44

 urč

-0.41

 ontvangen

-0.39

 daarvoor

-0.35

 eerder

-0.35

 zeichnen

-0.35

 opdracht

-0.33

zurück

-0.33

 geleden

-0.33

 actionMode

-0.32

POSITIVE LOGITS

tagHelperRunner

0.76

 fool

0.73

Cycle

0.71

 autorytatywna

0.71

 CYCLE

0.70

 myſelf

0.69

 Normal

0.69

 Cycle

0.68

 '\\;'

0.68

Normal

0.68

Activations Density 1.848%

references to deception, manipulation, and falsehoods

No Comments

No Known Activations