INDEX

Explanations

repeated references to singular concepts or entities associated with significant themes

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

 something

-0.14

[at

-0.14

 Lois

-0.14

 Alone

-0.14

zd

-0.13

 iets

-0.13

Ø²Ø©

-0.13

Ã´t

-0.13

ikan

-0.13

retch

-0.13

POSITIVE LOGITS

 that

0.32

 whose

0.29

whose

0.28

that

0.28

 ÏĢÎ¿Ïħ

0.27

 which

0.25

who

0.24

 mÃł

0.24

	that

0.23

 Ø§ÙĦØ°ÙĬ

0.23

Activations Density 0.295%

repeated references to singular concepts or entities associated with significant themes

No Comments

No Known Activations