INDEX

Explanations

references to secrecy and hidden information

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

fnlp/Llama3_1-8B-Base-LXM-8x/Llama3_1-8B-Base-L22M-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

imetype

-0.16

orest

-0.16

Ã¸r

-0.15

 Kelly

-0.15

andest

-0.14

.XPath

-0.14

Ã¼m

-0.14

 <!--[

-0.14

record

-0.14

Kelly

-0.14

POSITIVE LOGITS

 reve

0.17

era

0.16

ixel

0.16

reve

0.15

 Toll

0.15

elo

0.15

 release

0.15

andan

0.15

á»ĥn

0.15

bol

0.14

Activations Density 0.230%

references to secrecy and hidden information

No Comments

No Known Activations