INDEX

Explanations

references to returning to a previous or normal state

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

ledon

-0.15

Ã¼tÃ¼n

-0.15

 Various

-0.15

zw

-0.15

_pv

-0.14

 lleg

-0.14

onna

-0.14

ibur

-0.14

ahoma

-0.13

osten

-0.13

POSITIVE LOGITS

 normal

0.33

 basics

0.31

normal

0.27

-normal

0.25

 roots

0.25

 sender

0.25

roots

0.23

 Basics

0.23

 NORMAL

0.23

Normal

0.23

Activations Density 0.096%

references to returning to a previous or normal state

No Comments

No Known Activations