INDEX

Explanations

terms related to burning or destruction

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

 Fauc

-0.17

akt

-0.16

ve

-0.15

oo

-0.15

ibel

-0.15

ennen

-0.15

Bindable

-0.15

aden

-0.15

AW

-0.15

POSITIVE LOGITS

ished

0.28

 alive

0.28

 Ð´Ð¾ÑĤ

0.26

-toast

0.23

ISHED

0.22

 Alive

0.22

ishing

0.21

alive

0.21

 bridges

0.20

inate

0.20

Activations Density 0.030%

terms related to burning or destruction

No Comments

No Known Activations