INDEX

Explanations

phrases related to instructions or descriptions, possibly in a technical context

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

__(/*!

-0.81

 figliu

-0.80

 înal

-0.79

 stră

-0.72

 Hellen

-0.71

 desgracia

-0.71

 prisms

-0.71

 argint

-0.71

mereka

-0.70

 băr

-0.70

POSITIVE LOGITS

neurial

0.79

indeki

0.72

“

0.71

ンの

0.70

itní

0.69

ールの

0.69

‘

0.68

 대한

0.68

いた

0.67

ETIC

0.67

Activations Density 0.005%

phrases related to instructions or descriptions, possibly in a technical context

No Comments

No Known Activations