INDEX

Explanations

references to collaboration and assistance

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

avin

-0.17

/reference

-0.16

vil

-0.15

Ð¸ÑģÐ»Ð¾Ð²

-0.15

enti

-0.14

å·ŀ

-0.14

ÏĦÎŃ

-0.14

shade

-0.14

ponder

-0.14

otty

-0.14

POSITIVE LOGITS

 applicable

0.22

 appropriate

0.22

 needed

0.22

needed

0.21

 Needed

0.20

appropriate

0.19

 Necessary

0.18

Needed

0.18

icensed

0.17

 necessary

0.17

Activations Density 0.198%

references to collaboration and assistance

No Comments

No Known Activations