INDEX

Explanations

references to miraculous events or occurrences

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

onso

-0.16

illet

-0.15

mh

-0.15

emez

-0.15

owi

-0.15

marks

-0.14

ovo

-0.14

åı

-0.14

onen

-0.14

ties

-0.14

POSITIVE LOGITS

roring

0.32

iam

0.28

iams

0.27

rored

0.27

rors

0.26

acles

0.26

acle

0.25

ACLE

0.23

illis

0.21

rror

0.21

Activations Density 0.008%

references to miraculous events or occurrences

No Comments

No Known Activations