INDEX

Explanations

words associated with brightness or illumination

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

 blown

-0.23

 drawn

-0.19

 thrown

-0.18

 beaten

-0.18

 ridden

-0.17

venture

-0.17

thrown

-0.17

isation

-0.17

iland

-0.17

itation

-0.16

POSITIVE LOGITS

äºĨ

0.33

äºĨä¸Ģ

0.32

ed

0.29

 äºĨ

0.28

ified

0.25

ized

0.23

ied

0.22

ited

0.21

ened

0.21

åĩºäºĨ

0.21

Activations Density 0.106%

words associated with brightness or illumination

No Comments

No Known Activations