INDEX

Explanations

hierarchical structures or subsections in documents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

ones

-0.16

BP

-0.16

lek

-0.16

re

-0.15

le

-0.15

ct

-0.15

BC

-0.15

lo

-0.14

sto

-0.14

less

-0.14

POSITIVE LOGITS

inx

0.16

ix

0.15

zcze

0.14

ÙģØ§Øª

0.14

notif

0.14

/Foundation

0.14

imap

0.14

oki

0.14

ide

0.14

atre

0.14

Activations Density 0.105%

hierarchical structures or subsections in documents

No Comments

No Known Activations