INDEX

Explanations

negations or phrases indicating something has never happened or been achieved

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

"never" and related negations

np_acts-logits-general · gemini-2.0-flash

occurrence of never

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

 متحده

-0.62

AOC

-0.58

AOC

-0.56

 isSelected

-0.55

bufio

-0.55

Pyx

-0.52

 transfieras

-0.52

 OMITTED

-0.50

SqlClient

-0.50

IBOutlet

-0.50

POSITIVE LOGITS

 never

1.29

 NEVER

1.21

never

1.20

 Never

1.18

Never

1.15

NEVER

1.12

 ever

1.03

nunca

0.95

 EVER

0.94

ever

0.93

Activations Density 0.095%

negations or phrases indicating something has never happened or been achieved

"never" and related negations

occurrence of never

No Comments

No Known Activations