INDEX

Explanations

occurrences of manipulation or deception tactics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/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.21.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

 gesteld

-0.48

seamnă

-0.46

 nivelul

-0.46

 geïsole

-0.42

zeiro

-0.42

 inhoud

-0.41

iub

-0.41

 zijne

-0.40

 betrekking

-0.40

 gogh

-0.40

POSITIVE LOGITS

 trick

0.99

 tricks

0.92

 STRATEGY

0.88

 strategy

0.83

 Tricks

0.81

 Trick

0.80

 cunning

0.79

 STRATEG

0.78

trick

0.78

 strateg

0.78

Activations Density 0.533%

occurrences of manipulation or deception tactics

No Comments

No Known Activations