INDEX

Explanations

variations of the prefix "en-," likely indicating a focus on verbs or actions related to becoming or causing to be

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

Ïİ

-0.20

onya

-0.17

tar

-0.17

tips

-0.16

town

-0.16

itionally

-0.15

tg

-0.15

tul

-0.15

ings

-0.15

tz

-0.15

POSITIVE LOGITS

sembles

0.32

igmatic

0.24

grav

0.23

ceph

0.23

visions

0.22

shr

0.22

igma

0.22

.wikipedia

0.21

forces

0.21

vious

0.21

Activations Density 0.013%

variations of the prefix "en-," likely indicating a focus on verbs or actions related to becoming or causing to be

No Comments

No Known Activations