INDEX

Explanations

references to the name "Norm" and its variants in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

assen

-0.17

ion

-0.16

roller

-0.15

ioned

-0.15

hips

-0.15

stry

-0.15

æ¢

-0.15

NT

-0.15

erase

-0.14

 Clem

-0.14

POSITIVE LOGITS

ative

0.28

andy

0.28

ATIVE

0.22

alement

0.21

atively

0.20

ativa

0.20

anton

0.19

linger

0.18

ativity

0.18

aller

0.17

Activations Density 0.012%

references to the name "Norm" and its variants in various contexts

No Comments

No Known Activations