INDEX

Explanations

punctuation marks, specifically commas and colons

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

open-r1/OpenR1-Math-220k

Features

32,768

Data Type

float32

Hook Name

blocks.15.hook_resid_post

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Embeds

IFrame

Link

Negative Logits

piler

-0.07

Ä¢

-0.07

ziej

-0.07

inia

-0.06

ãĥ¼ãĥĦ

-0.06

orrent

-0.06

$MESS

-0.06

venta

-0.06

urga

-0.06

_ABORT

-0.06

POSITIVE LOGITS

ember

0.07

 nackte

0.07

 chatte

0.06

 unlimited

0.06

ere

0.06

guy

0.06

oval

0.06

huh

0.06

Activations Density 0.048%

No Known Activations