INDEX

Explanations

numerical values and specific formatting characters

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

teen

-0.21

een

-0.21

-0.18

eenth

-0.18

-0.17

XIV

-0.17

XV

-0.17

-0.16

XVI

-0.16

POSITIVE LOGITS

0.51

0.45

 nine

0.43

 ninth

0.40

nine

0.39

 eight

0.38

ä¹Ŀ

0.36

 Nine

0.35

 Ninth

0.35

 eighth

0.33

Activations Density 0.152%

numerical values and specific formatting characters

No Comments

No Known Activations