INDEX

Explanations

phrases that express direct addresses or inclusivity towards the audience

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

utor

-0.15

äººäºº

-0.14

rts

-0.14

_SPE

-0.14

 secondary

-0.13

ä¼ı

-0.13

 Surveillance

-0.13

$(

-0.13

oci

-0.13

 stride

-0.13

POSITIVE LOGITS

 readers

0.31

 reader

0.26

 Readers

0.25

 unfamiliar

0.24

reader

0.22

 Reader

0.22

Reader

0.21

 reading

0.20

-reader

0.20

 wondering

0.19

Activations Density 0.071%

phrases that express direct addresses or inclusivity towards the audience

No Comments

No Known Activations