INDEX

Explanations

terms related to harassment, particularly in social and cultural contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

iard

-0.22

gear

-0.20

ãģ¨ãģĵãĤį

-0.19

dle

-0.18

etary

-0.18

ular

-0.17

 Fargo

-0.16

æ¨£

-0.16

æł·çļĦ

-0.16

going

-0.16

POSITIVE LOGITS

mony

0.20

red

0.19

rr

0.19

oon

0.18

cÄ±

0.18

bitrary

0.17

iously

0.17

itect

0.16

oons

0.16

ãģ¹ãģį

0.16

Activations Density 0.119%

terms related to harassment, particularly in social and cultural contexts

No Comments

No Known Activations