INDEX

Explanations

references to crime and criminal activities

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

 Super

-0.19

SUPER

-0.19

 SUPER

-0.18

 super

-0.18

Super

-0.17

super

-0.17

oka

-0.16

uper

-0.16

(super

-0.16

 hero

-0.15

POSITIVE LOGITS

tut

0.15

 Ð³Ð»Ñĥ

0.15

opsy

0.15

ãĤ¶ãĥ¼

0.14

ools

0.14

atsu

0.14

aupt

0.13

ÑĤÐ¾Ñĩ

0.13

à¹ģà¸«à¸¥

0.13

iete

0.13

Activations Density 0.073%

references to crime and criminal activities

No Comments

No Known Activations