INDEX

Explanations

code documentation summaries and parameter annotations

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

angu

-0.19

ÂŃt

-0.17

uda

-0.17

zip

-0.15

.pojo

-0.15

 Furn

-0.14

chan

-0.14

 Jackson

-0.14

iqu

-0.14

inar

-0.14

POSITIVE LOGITS

 cref

0.17

fully

0.16

otics

0.15

metic

0.15

ToFit

0.15

rÃ¥

0.14

_IW

0.14

atti

0.14

>↵

0.13

ByExample

0.13

Activations Density 0.003%

code documentation summaries and parameter annotations

No Comments

No Known Activations