INDEX

Explanations

references to software or technology features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

uze

-0.17

åĪ·

-0.14

 Range

-0.14

.Host

-0.14

hazi

-0.14

orry

-0.13

ialog

-0.13

çĪ

-0.13

 ranges

-0.13

('.'

-0.13

POSITIVE LOGITS

 Tiles

0.24

 Metro

0.22

 tiles

0.21

Metro

0.20

dde

0.20

 tile

0.20

 Start

0.20

 Dock

0.19

 metro

0.19

 desktop

0.19

Activations Density 0.047%

references to software or technology features

No Comments

No Known Activations