INDEX

Explanations

punctuation marks and elements related to formatting in text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.hook_resid_post

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

ela

-0.16

andas

-0.15

as

-0.15

ÅĻ

-0.15

 geschichten

-0.14

agem

-0.14

era

-0.14

é¼

-0.14

chal

-0.13

atan

-0.13

POSITIVE LOGITS

am

0.28

maj

0.28

ami

0.28

 mint

0.24

mik

0.22

ez

0.22

amik

0.21

 erre

0.21

ah

0.21

hol

0.20

Activations Density 0.000%

punctuation marks and elements related to formatting in text

No Comments

No Known Activations