INDEX

Explanations

references to sacrifice or harm involving religious figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

ÎµÏħ

-0.13

rovers

-0.13

MMdd

-0.13

DIRECTORY

-0.13

olar

-0.13

Mui

-0.12

innie

-0.12

oya

-0.12

POSITIVE LOGITS

ONS

0.17

 åĵģ

0.16

enthal

0.15

.asp

0.15

illy

0.15

iskey

0.14

arshal

0.14

Ze

0.14

azole

0.14

Activations Density 0.011%

references to sacrifice or harm involving religious figures

No Comments

No Known Activations