INDEX

Explanations

terms related to racial and religious extremism

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

akis

-0.18

rrha

-0.17

 nackt

-0.16

ffect

-0.15

vale

-0.15

icamente

-0.15

gay

-0.14

ilde

-0.14

Gay

-0.14

 láº¡

-0.14

POSITIVE LOGITS

 plac

0.16

Ves

0.15

 Borders

0.15

·

0.15

å¹

0.15

 Caul

0.14

 host

0.14

ç©į

0.14

aus

0.14

/tos

0.14

Activations Density 0.185%

terms related to racial and religious extremism

No Comments

No Known Activations