INDEX

Explanations

references to white identity and supremacy

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

 Bucc

-0.17

ORK

-0.16

 blacks

-0.16

ÐµÑĢÐ¸Ð½

-0.15

InstanceOf

-0.15

pcs

-0.14

ibel

-0.14

geme

-0.14

ypes

-0.13

worthy

-0.13

POSITIVE LOGITS

 supremacy

0.23

 privilege

0.23

 male

0.23

-owned

0.22

 suprem

0.21

 supremacist

0.21

bread

0.20

 males

0.20

-collar

0.20

Male

0.19

Activations Density 0.021%

references to white identity and supremacy

No Comments

No Known Activations