INDEX

Explanations

references to a history of accountability or achievement

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

oplast

-0.15

.esp

-0.15

selectAll

-0.15

 rewind

-0.15

gnu

-0.14

opus

-0.14

anol

-0.14

 spread

-0.14

cy

-0.14

POSITIVE LOGITS

strand

0.20

à¹ĭ

0.17

 DISPATCH

0.15

tro

0.15

eted

0.15

inson

0.15

 GOODMAN

0.15

 Ð¡ÑĥÐ´

0.15

 DISCLAIM

0.14

bow

0.14

Activations Density 0.021%

references to a history of accountability or achievement

No Comments

No Known Activations