INDEX

Explanations

incidents involving tragic events or accidents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

oro

-0.15

 Wander

-0.14

cpt

-0.14

.EventQueue

-0.14

eton

-0.14

ianne

-0.14

oca

-0.14

 appropri

-0.14

lox

-0.14

igel

-0.13

POSITIVE LOGITS

 preparations

0.20

 preparing

0.20

 returning

0.20

 innoc

0.19

 routine

0.19

 attending

0.18

åĩĨå¤ĩ

0.17

routine

0.17

æŃ£åľ¨

0.17

 Ð¾ÑĩÐµÑĢÐµÐ´

0.17

Activations Density 0.324%

incidents involving tragic events or accidents

No Comments

No Known Activations