INDEX

Explanations

phrases that indicate varying degrees of magnitude or intensity, particularly using qualifiers like "very" and "high."

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

Negative Logits

Re

-0.64

re

-0.60

Re

-0.54

par

-0.53

En

-0.53

und

-0.52

RE

-0.52

-0.49

Ch

-0.48

POSITIVE LOGITS

 itſelf

0.97

 MainAxisSize

0.90

GEBURTSDATUM

0.90

 uſed

0.90

 ſhould

0.89

MessageTagHelper

0.85

ItemBackground

0.85

 ſtill

0.84

 fhould

0.84

новниш

0.83

Activations Density 0.092%

phrases that indicate varying degrees of magnitude or intensity, particularly using qualifiers like "very" and "high."

No Known Activations