INDEX

Explanations

references to specific awards or titles, particularly those abbreviated with an "L" followed by a number

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

arsers

-0.15

mant

-0.15

asty

-0.15

fo

-0.15

asad

-0.15

iesen

-0.14

 mant

-0.14

 alike

-0.14

POSITIVE LOGITS

adb

0.27

ough

0.24

land

0.24

lando

0.23

args

0.23

lan

0.21

lang

0.21

iscard

0.21

enton

0.20

lund

0.20

Activations Density 0.010%

references to specific awards or titles, particularly those abbreviated with an "L" followed by a number

No Comments

No Known Activations