INDEX

Explanations

words related to security and risk factors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_22/width_16k/average_l0_123

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

ubro

-0.46

 Inde

-0.45

 propa

-0.44

pons

-0.43

 repress

-0.42

Dee

-0.42

Jos

-0.42

 Савезне

-0.42

Cyfeiriadau

-0.41

 Tira

-0.41

POSITIVE LOGITS

 paille

0.50

Hentet

0.50

aarrggbb

0.48

 vueltas

0.45

 transición

0.45

 recuperación

0.45

UnitTesting

0.44

protetor

0.44

intenant

0.44

 AssemblyTitle

0.43

Activations Density 0.558%

words related to security and risk factors

No Comments

No Known Activations