INDEX

Explanations

questions beginning with "What."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

 whats

-0.17

itsu

-0.15

 Whats

-0.14

Å¡e

-0.14

eks

-0.14

amework

-0.14

ÑĸÑģ

-0.13

é³

-0.13

yen

-0.13

æ¡Ī

-0.13

POSITIVE LOGITS

are

0.32

 does

0.28

Are

0.27

_are

0.26

do

0.25

 Does

0.23

æĺ¯æĪĳ

0.22

.are

0.21

Are

0.20

 exactly

0.19

Activations Density 0.068%

questions beginning with "What."

No Comments

No Known Activations