INDEX

Explanations

references to firearms and gun-related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

ÐµÑĢÐ¸Ð½

-0.18

hra

-0.17

HORT

-0.16

©

-0.15

cue

-0.15

ØªÙĨ

-0.15

 bombing

-0.15

merc

-0.15

cloth

-0.15

acus

-0.14

POSITIVE LOGITS

pow

0.30

ned

0.22

ning

0.22

shots

0.20

metal

0.20

ny

0.20

linger

0.19

erals

0.18

ners

0.18

smith

0.18

Activations Density 0.023%

references to firearms and gun-related terminology

No Comments

No Known Activations