INDEX

Explanations

suggestions and proposals within the text

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

ucha

-0.18

old

-0.15

adar

-0.15

ilde

-0.15

readcr

-0.15

adge

-0.14

-Ð·Ð°

-0.14

occo

-0.14

atts

-0.14

à§į

-0.14

POSITIVE LOGITS

ively

0.34

ive

0.26

 strongly

0.19

/request

0.19

entially

0.18

IVE

0.18

ìĤ¬íķŃ

0.18

ibility

0.17

 ìĤ¬íķŃ

0.17

 ways

0.17

Activations Density 0.021%

suggestions and proposals within the text

No Comments

No Known Activations