INDEX

Explanations

calls to action or invitations to engage further with content

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

minate

-0.15

linger

-0.14

idor

-0.14

ALA

-0.14

dos

-0.14

Probe

-0.13

grÃ©

-0.13

Ð¾ÑĢÐ¾ÑĪ

-0.13

anda

-0.13

 thin

-0.13

POSITIVE LOGITS

 Morton

0.15

ovsky

0.15

indre

0.14

âĨĴâĨĴ

0.13

ç´

0.13

hips

0.13

ouz

0.13

 âĨĴ↵↵

0.13

UID

0.13

 Clyde

0.13

Activations Density 0.041%

calls to action or invitations to engage further with content

No Comments

No Known Activations