INDEX

Explanations

phrases indicating quantity or multiplicity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

aries

-0.16

enne

-0.16

ALSE

-0.16

istrovstvÃŃ

-0.15

pedia

-0.14

ADO

-0.14

eld

-0.14

OLDER

-0.14

jen

-0.14

isms

-0.13

POSITIVE LOGITS

Tus

0.14

ourcem

0.14

/etc

0.14

iselect

0.14

awk

0.13

ouro

0.13

assword

0.13

 Concord

0.13

Fay

0.13

æµ´

0.13

Activations Density 0.040%

phrases indicating quantity or multiplicity

No Comments

No Known Activations