INDEX

Explanations

phrases related to authorship and publication context

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

-0.16

 just

-0.16

Rig

-0.14

 Prel

-0.14

á»ī

-0.14

hor

-0.14

isu

-0.14

 anywhere

-0.14

hor

-0.14

POSITIVE LOGITS

 originally

0.23

Originally

0.23

 original

0.23

 Originally

0.21

/original

0.19

(original

0.19

åİŁ

0.18

original

0.18

 Ð¾ÑĢÐ¸Ð³

0.18

 nguyÃªn

0.17

Activations Density 0.109%

phrases related to authorship and publication context

No Comments

No Known Activations