INDEX

Explanations

numerical values or statistics related to addresses and quantities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

dÃ¼r

-0.17

vell

-0.17

ÑĴ

-0.16

áº»

-0.16

iyim

-0.15

Å¾e

-0.15

ano

-0.14

atre

-0.14

bdb

-0.14

osoph

-0.14

POSITIVE LOGITS

å¹´ä»£

0.20

odd

0.17

ish

0.16

ial

0.16

ies

0.15

wick

0.15

ians

0.15

allon

0.15

shire

0.15

Activations Density 0.131%

numerical values or statistics related to addresses and quantities

No Comments

No Known Activations