INDEX

Explanations

phrases related to challenges and issues faced 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 @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

benchmark

-0.16

ibrated

-0.15

 discrepancy

-0.15

ients

-0.15

"<?

-0.14

åı

-0.14

mej

-0.13

 DÃŃky

-0.13

ndo

-0.13

getBlock

-0.13

POSITIVE LOGITS

 challenges

0.24

 unique

0.22

 serious

0.22

 risks

0.21

 privacy

0.20

unique

0.19

 practical

0.18

 safety

0.17

 inherent

0.17

 dile

0.17

Activations Density 0.125%

phrases related to challenges and issues faced in various contexts

No Comments

No Known Activations