INDEX

Explanations

terms related to racial or social justice issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

ammen

-0.15

 defaultCenter

-0.14

hq

-0.14

avr

-0.14

 Strip

-0.14

ipay

-0.13

ÑĢÑĸÐ¿

-0.13

alars

-0.13

IGH

-0.13

 Gaga

-0.13

POSITIVE LOGITS

 race

0.31

 quotas

0.28

 affirmative

0.28

race

0.27

 Race

0.27

 admissions

0.26

Race

0.26

 diversity

0.24

 quota

0.24

 racial

0.24

Activations Density 0.067%

terms related to racial or social justice issues

No Comments

No Known Activations