INDEX

Explanations

structured data formats, particularly nested arrays and objects

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

vat

-0.16

oth

-0.16

/pub

-0.15

ycin

-0.15

nings

-0.15

ARRY

-0.15

vinces

-0.15

neau

-0.14

arry

-0.14

pon

-0.14

POSITIVE LOGITS

egers

0.15

urs

0.15

ennen

0.15

embros

0.15

PictureBox

0.15

ayment

0.14

idders

0.14

arker

0.14

erview

0.14

onym

0.14

Activations Density 0.037%

structured data formats, particularly nested arrays and objects

No Comments

No Known Activations