INDEX

Explanations

references to advertisements and related marketing content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

ekk

-0.18

uctions

-0.17

uction

-0.16

ecies

-0.16

ivot

-0.16

eczy

-0.15

embre

-0.15

eve

-0.15

uco

-0.15

dling

-0.15

POSITIVE LOGITS

rien

0.27

olph

0.27

ria

0.26

el

0.25

nan

0.24

kins

0.23

ity

0.23

air

0.22

riel

0.22

eline

0.20

Activations Density 0.013%

references to advertisements and related marketing content

No Comments

No Known Activations