INDEX

Explanations

references to specific individuals or characters denoted by the letter 'A'

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

linger

-0.18

xes

-0.15

oler

-0.15

uld

-0.15

ingles

-0.15

å¤Ħ

-0.15

ux

-0.15

/react

-0.15

opp

-0.15

ubl

-0.15

POSITIVE LOGITS

ish

0.23

apk

0.23

nees

0.23

ryan

0.22

stit

0.21

ham

0.21

am

0.20

ash

0.20

div

0.20

asma

0.20

Activations Density 0.040%

references to specific individuals or characters denoted by the letter 'A'

No Comments

No Known Activations