INDEX

Explanations

phrases indicating purpose or intention

oai_token-act-pair · gpt-4o-mini

New Auto-Interp

Configuration

fnlp/Llama-Scope-R1-Distill/800M-Slimpajama-0-OpenR1-Math-220k/L15R

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Hzfinfdu/SlimPajama-3B

Features

32,768

Data Type

float32

Hook Name

blocks.15.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

muz

-0.14

ESA

-0.14

StackNavigator

-0.14

than

-0.13

ãĥ©ãĤ¹

-0.13

 reve

-0.13

stell

-0.13

luet

-0.13

ritis

-0.13

dit

-0.13

POSITIVE LOGITS

yz

0.15

ibu

0.14

ilers

0.13

 plÃ¡

0.13

asting

0.13

.poster

0.13

.getContentPane

0.13

ivan

0.13

ielding

0.13

rame

0.12

Activations Density 0.217%

phrases indicating purpose or intention

No Comments

No Known Activations