INDEX

Explanations

concepts related to deceptive practices or misleading tactics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 23-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_23/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.23.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

="#"><

-0.30

WriteLiteral

-0.28

=""><

-0.28

Bai

-0.27

tei

-0.27

rbp

-0.27

Michigan

-0.27

 Michigan

-0.26

lesias

-0.26

 samt

-0.26

POSITIVE LOGITS

verwijspagina

0.85

featureID

0.72

rungsseite

0.68

клопе

0.66

aarrggbb

0.65

 HttpNotFound

0.62

nextProps

0.62

DockStyle

0.61

didSet

0.60

+#+

0.60

Activations Density 1.179%

concepts related to deceptive practices or misleading tactics

No Comments

No Known Activations