INDEX

Explanations

command phrases or instructions related to an interface

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

à¤¾à¤¸à¤¨

-0.16

gili

-0.15

odyn

-0.15

anken

-0.15

/buttons

-0.15

Dia

-0.14

specifier

-0.14

isy

-0.14

Î²Î¬Î»

-0.13

 GenerationType

-0.13

POSITIVE LOGITS

 select

0.29

 click

0.26

select

0.24

 expand

0.24

 hover

0.23

 locate

0.23

 navigate

0.22

 scroll

0.21

hit

0.21

 selects

0.21

Activations Density 0.063%

command phrases or instructions related to an interface

No Comments

No Known Activations