INDEX

Explanations

phrases that highlight moral or ethical dilemmas related to harm and responsibility

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

dera

-0.20

anki

-0.16

enson

-0.15

ifornia

-0.14

 curtains

-0.14

cke

-0.13

arsing

-0.13

ocket

-0.13

aud

-0.13

POSITIVE LOGITS

nor

0.23

 anymore

0.20

ANY

0.16

nor

0.15

 anybody

0.14

licht

0.14

idia

0.14

 newPosition

0.14

Î½Î¿Ïį

0.14

any

0.14

Activations Density 0.213%

phrases that highlight moral or ethical dilemmas related to harm and responsibility

No Comments

No Known Activations