INDEX

Explanations

words related to security, confidentiality, and ethical practices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

Ĺ

-0.14

ÏģÎ¸

-0.14

Rap

-0.14

lai

-0.14

ÛĮØ±ÛĮ

-0.13

å¢¨

-0.13

 Prospect

-0.13

 censor

-0.13

ment

-0.13

 callable

-0.13

POSITIVE LOGITS

abal

0.16

/umd

0.15

utas

0.15

ldr

0.15

ä¸Ķ

0.14

obel

0.14

.githubusercontent

0.14

ç©´

0.14

 Ä°mpar

0.14

aucoup

0.14

Activations Density 0.371%

words related to security, confidentiality, and ethical practices

No Comments

No Known Activations