INDEX

Explanations

negative or dismissive phrases

oai_token-act-pair · gpt-4o-mini

New Auto-Interp

Configuration

chanind/gemma-2-2b-batch-topk-matryoshka-saes-w-32k-l0-40/standard/blocks.12.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Architecture

jumprelu

Context Size

1,024

Dataset

chanind/pile-uncopyrighted-gemma-1024-abbrv-1B

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

IONI

-0.64

 culturelles

-0.64

 vocês

-0.63

Bibliograf

-0.63

 SALUD

-0.62

 ustedes

-0.61

сього

-0.60

Torres

-0.60

 suivantes

-0.60

 což

-0.60

POSITIVE LOGITS

no

1.73

No

1.53

NO

1.41

No

1.37

no

1.12

NO

1.11

 Noyes

1.06

Нет

1.00

igno

0.98

nof

0.97

Activations Density 0.523%

negative or dismissive phrases

No Comments

No Known Activations