INDEX

Explanations

references to individuals and their formal titles

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

 levens

-0.48

ințe

-0.41

 jalur

-0.39

ина

-0.38

aringan

-0.37

enlace

-0.36

 inşa

-0.35

 zicht

-0.35

 ogrodow

-0.35

 cerrada

-0.34

POSITIVE LOGITS

Mr

1.04

Mr

0.98

rungsseite

0.92

 Monfieur

0.89

 Messrs

0.84

mr

0.83

 آقای

0.82

principalColumn

0.80

Mrs

0.80

 Mister

0.77

Activations Density 0.163%

references to individuals and their formal titles

No Comments

No Known Activations