INDEX

Explanations

phrases indicating an unsettling or abnormal situation

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

Negative Logits

ramer

-0.15

emer

-0.14

èįĴ

-0.14

mps

-0.13

 experiences

-0.13

 temptation

-0.13

fait

-0.13

ska

-0.13

 khÃ³

-0.13

éŀ

-0.13

POSITIVE LOGITS

 wrong

0.33

wrong

0.30

 Wrong

0.27

 WRONG

0.25

Wrong

0.25

 fish

0.23

fish

0.22

_wrong

0.22

 Fish

0.18

Fish

0.18

Activations Density 0.069%

phrases indicating an unsettling or abnormal situation

No Known Activations