INDEX

Explanations

conditional phrases and expressions of potential actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

rp

-0.15

utow

-0.15

ters

-0.15

 Laud

-0.14

Ú©Øª

-0.14

 JFactory

-0.14

zig

-0.14

ÐµÐ»ÐµÐ¼

-0.14

seealso

-0.14

WISE

-0.14

POSITIVE LOGITS

so

0.21

 excuse

0.20

 pardon

0.16

ardon

0.16

iw

0.16

 shall

0.15

 excuses

0.15

ivirus

0.14

Exc

0.14

éĤ£ä¹Ī

0.14

Activations Density 0.023%

conditional phrases and expressions of potential actions

No Comments

No Known Activations