INDEX

Explanations

occurrences of the word "cov" or its variations, indicating a focus on the theme related to "coverage" or "COVID"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

erap

-0.17

erro

-0.16

å¹ķ

-0.16

ustum

-0.15

eron

-0.15

Ã³l

-0.15

alÄ±

-0.15

ÑĥÐ»Ð¸

-0.15

esini

-0.15

erne

-0.15

POSITIVE LOGITS

ington

0.25

INGTON

0.23

ariate

0.21

etailed

0.20

ariance

0.20

iedo

0.19

cov

0.17

ariant

0.17

Ã¡bb

0.17

aleur

0.15

Activations Density 0.014%

occurrences of the word "cov" or its variations, indicating a focus on the theme related to "coverage" or "COVID"

No Comments

No Known Activations