INDEX

Explanations

references to safety, security, and conflict in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

ibu

-0.15

Î¯ÏĦ

-0.14

ipp

-0.14

 Beaut

-0.14

 basically

-0.14

ãĤ¦ãĤ¹

-0.14

Fu

-0.13

SITE

-0.13

_DEFINE

-0.13

agen

-0.13

POSITIVE LOGITS

 Ð¿Ð¾ÐºÐ°

0.21

 until

0.19

 zatÃŃm

0.19

until

0.18

 presently

0.17

 till

0.17

inkel

0.17

 Until

0.17

 initially

0.16

 currently

0.16

Activations Density 0.222%

references to safety, security, and conflict in various contexts

No Comments

No Known Activations