INDEX

Explanations

phrases referencing concepts of design, moral philosophy, and hierarchical relationships

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

ued

-0.16

amp

-0.16

arium

-0.15

inke

-0.14

zeichnet

-0.14

ude

-0.13

ing

-0.13

pt

-0.13

Ã¡o

-0.13

omaly

-0.13

POSITIVE LOGITS

 sheer

0.23

 matters

0.17

 finances

0.16

how

0.16

 specifically

0.15

 personal

0.15

 konkrÃ©t

0.15

 general

0.14

Sharper

0.14

tual

0.14

Activations Density 0.122%

phrases referencing concepts of design, moral philosophy, and hierarchical relationships

No Comments

No Known Activations