INDEX

Explanations

references to missions or objectives, often in a formal or official context

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

OFF

-0.15

ãĥĨãĥ«

-0.15

ture

-0.15

 áº¤

-0.15

lee

-0.15

ddy

-0.15

iman

-0.14

ish

-0.14

/she

-0.14

POSITIVE LOGITS

naire

0.19

naires

0.19

aries

0.18

WARDED

0.17

aira

0.16

omez

0.16

ColorBrush

0.16

éĢĶ

0.15

/target

0.15

avicon

0.15

Activations Density 0.013%

references to missions or objectives, often in a formal or official context

No Comments

No Known Activations