INDEX

Explanations

negation or phrases indicating refusal or doubt

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

vÄĽt

-0.15

inci

-0.15

******/

-0.14

áº£y

-0.14

elay

-0.14

ç¸

-0.13

itals

-0.13

ock

-0.13

stration

-0.13

 RootState

-0.13

POSITIVE LOGITS

 necessarily

0.38

 sure

0.28

withstanding

0.25

 least

0.23

 because

0.23

least

0.23

sure

0.23

 counting

0.22

 Sure

0.21

 vice

0.21

Activations Density 0.079%

negation or phrases indicating refusal or doubt

No Comments

No Known Activations