INDEX

Explanations

interrogative structures and elements related to inquiry

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

âĢŀP

-0.15

uat

-0.15

 wenn

-0.15

 nameof

-0.14

klu

-0.14

emailer

-0.14

asto

-0.14

igger

-0.14

if

-0.14

 >č↵

-0.13

POSITIVE LOGITS

 consider

0.33

 here

0.31

why

0.28

consider

0.28

 Consider

0.28

you

0.27

Consider

0.26

 perhaps

0.25

 head

0.25

try

0.25

Activations Density 0.166%

interrogative structures and elements related to inquiry

No Comments

No Known Activations

interrogative structures and elements related to inquiry

No Comments

No Known Activations