INDEX

Explanations

references to users and user-related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

UserInfo

-0.23

UserProfile

-0.23

UserData

-0.22

User

-0.21

UserRole

-0.20

ses

-0.18

_users

-0.17

UserService

-0.17

si

-0.17

Info

-0.17

POSITIVE LOGITS

names

0.22

 Datagram

0.22

-Agent

0.21

-defined

0.20

InteractionEnabled

0.20

fully

0.19

itarian

0.19

ful

0.18

-friendly

0.18

fare

0.18

Activations Density 0.047%

references to users and user-related concepts

No Comments

No Known Activations