INDEX

Explanations

references to various parts and sections of a document or object

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

"of the"

np_acts-logits-general · gemini-2.0-flash

of the

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

paravant

-0.39

sweise

-0.36

 åren

-0.35

 nahilalakip

-0.34

 состоя

-0.33

 varandra

-0.33

 erbjud

-0.33

 include

-0.32

Dış

-0.32

 nabíz

-0.32

POSITIVE LOGITS

AtEndOf

0.75

EndOf

0.72

endOf

0.68

enfone

0.61

 Мексичка

0.60

WriteTagHelper

0.60

Inside

0.58

siri

0.57

astie

0.56

zweig

0.56

Activations Density 0.377%

references to various parts and sections of a document or object

"of the"

of the

No Comments

No Known Activations