INDEX

Explanations

phrases that communicate identity or the nature of subjects

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

obody

-0.14

ekyll

-0.14

inton

-0.14

à¥Ĥà¤ļ

-0.13

acco

-0.13

ieder

-0.13

 Äįast

-0.13

ÏĥÏĥ

-0.13

ether

-0.12

POSITIVE LOGITS

 having

0.49

 being

0.48

being

0.41

having

0.38

Having

0.34

 Having

0.34

 ayant

0.33

 sendo

0.32

 Being

0.31

Being

0.30

Activations Density 0.122%

phrases that communicate identity or the nature of subjects

No Comments

No Known Activations