INDEX

Explanations

instances of the word "over" and its variations in context to quantities or frequency

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

 overview

-0.23

overview

-0.22

 Overview

-0.21

Overview

-0.21

 overloaded

-0.21

overwrite

-0.18

-overlay

-0.18

 overnight

-0.18

over

-0.18

 overlap

-0.18

POSITIVE LOGITS

alls

0.29

lying

0.28

tones

0.27

lord

0.27

tures

0.26

heard

0.25

flows

0.25

hang

0.25

ture

0.24

drive

0.24

Activations Density 0.152%

instances of the word "over" and its variations in context to quantities or frequency

No Comments

No Known Activations