INDEX

Explanations

the word "nine" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

Ð¾Ð¼ÐµÑĢ

-0.16

adla

-0.16

 DeepCopy

-0.15

ãĤ¤ãĤ¯

-0.15

cratch

-0.15

Zus

-0.14

acers

-0.14

.synthetic

-0.14

LATED

-0.14

ei

-0.14

POSITIVE LOGITS

ety

0.29

ete

0.28

eties

0.28

jas

0.27

ETY

0.26

otch

0.21

jac

0.20

jab

0.20

teen

0.19

hydr

0.18

Activations Density 0.006%

the word "nine" and its variations

No Comments

No Known Activations