INDEX

Explanations

references to selection processes or the act of choosing the right options

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

ÏģÎ¹ÏĥÏĦ

-0.16

à¸²à¸ķà¸£

-0.16

ESP

-0.15

myp

-0.14

BOSE

-0.14

ombs

-0.14

.IsAny

-0.13

ampus

-0.13

outers

-0.13

ronym

-0.13

POSITIVE LOGITS

 correct

0.60

 appropriate

0.50

 proper

0.47

correct

0.45

 wrong

0.44

æŃ£ç¡®

0.42

appropriate

0.41

 right

0.39

 Correct

0.38

Correct

0.37

Activations Density 0.094%

references to selection processes or the act of choosing the right options

No Comments

No Known Activations