INDEX

Explanations

references to race, specifically the term "white" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

edin

-0.20

ments

-0.19

ermint

-0.17

edImage

-0.17

nts

-0.17

nt

-0.17

ment

-0.16

ement

-0.16

rito

-0.15

Kov

-0.15

POSITIVE LOGITS

-collar

0.33

caps

0.31

hall

0.31

 supremacist

0.30

paper

0.29

board

0.29

bread

0.28

papers

0.27

fish

0.27

-hot

0.27

Activations Density 0.030%

references to race, specifically the term "white" and its variations

No Comments

No Known Activations