INDEX

Explanations

references to background elements or settings in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

åĵģ

-0.16

its

-0.15

apa

-0.15

ardi

-0.15

ep

-0.15

baum

-0.14

ego

-0.14

uyen

-0.14

igli

-0.14

ewan

-0.14

POSITIVE LOGITS

/background

0.22

/back

0.20

(background

0.19

 background

0.17

(back

0.17

.Background

0.15

APER

0.15

ultz

0.15

kus

0.15

 èĥ

0.15

Activations Density 0.022%

references to background elements or settings in various contexts

No Comments

No Known Activations