INDEX

Explanations

terms related to damage and brokenness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.28.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

rÃ³

-0.14

leta

-0.14

opensource

-0.14

ially

-0.14

ropa

-0.14

isd

-0.14

 Äįist

-0.14

NotFoundException

-0.13

abcdefghijklmnop

-0.13

GF

-0.13

POSITIVE LOGITS

æİī

0.23

ness

0.23

/loose

0.23

/dist

0.22

/problem

0.22

/null

0.21

/question

0.21

 condition

0.20

/conf

0.20

-down

0.20

Activations Density 0.209%

terms related to damage and brokenness

No Comments

No Known Activations