INDEX

Explanations

references to glass-related items or materials

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

stery

-0.16

icontrol

-0.16

ertools

-0.16

ved

-0.15

Ð¾ÑĪ

-0.15

Ã¡c

-0.15

east

-0.14

uche

-0.14

vfs

-0.14

memberof

-0.14

POSITIVE LOGITS

gow

0.30

(es

0.27

ses

0.26

ware

0.26

boro

0.23

ine

0.22

cock

0.22

work

0.22

works

0.21

Activations Density 0.011%

references to glass-related items or materials

No Comments

No Known Activations

references to glass-related items or materials

No Comments

No Known Activations