INDEX

Explanations

negations or phrases indicating something is not true or valid

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

"not" followed by an adjective

np_acts-logits-general · gemini-2.0-flash

not followed by word

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 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

setVerticalGroup

-0.94

 EconPapers

-0.93

 chi̍t

-0.92

 kaarangay

-0.91

contentLoaded

-0.89

 متعلقه

-0.88

WithIOException

-0.85

PreferredItem

-0.85

 ModelExpression

-0.84

 lenker

-0.82

POSITIVE LOGITS

 uncommon

0.70

bad

0.66

 fair

0.61

 unusual

0.53

 surprising

0.52

new

0.52

 meant

0.51

vium

0.51

 necessarily

0.48

bad

0.48

Activations Density 0.196%

negations or phrases indicating something is not true or valid

"not" followed by an adjective

not followed by word

No Comments

No Known Activations