INDEX

Explanations

language related to authority, rules, and compliance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

 yourself

-0.17

Ñľ

-0.17

ello

-0.16

 oneself

-0.16

 Yourself

-0.15

 myself

-0.14

 ApiController

-0.14

leyen

-0.14

æĶ¿åºľ

-0.14

jr

-0.14

POSITIVE LOGITS

us

0.20

 itself

0.20

iddi

0.17

itivity

0.16

iddle

0.15

its

0.15

acas

0.14

.promise

0.14

iten

0.14

alars

0.13

Activations Density 0.559%

language related to authority, rules, and compliance

No Comments

No Known Activations