INDEX

Explanations

references to functionality and operational issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

 nhÃ©

-0.16

abbo

-0.15

ventus

-0.14

inden

-0.14

")!=

-0.14

ä¹ī

-0.13

à¤§à¤°

-0.13

iddi

-0.13

chez

-0.13

Orm

-0.13

POSITIVE LOGITS

 properly

0.41

proper

0.39

due

0.35

 Proper

0.34

 proper

0.32

due

0.31

 sometimes

0.29

 correctly

0.27

 adequate

0.26

 certain

0.25

Activations Density 0.337%

references to functionality and operational issues

No Comments

No Known Activations