INDEX

Explanations

references to the Netherlands and Dutch entities

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

 meiden

-0.15

Zug

-0.15

 eskort

-0.15

.sf

-0.15

pev

-0.14

 dikke

-0.14

ÙĪÙĦÛĮ

-0.14

cek

-0.14

cheid

-0.14

ochond

-0.14

POSITIVE LOGITS

 Dutch

0.30

.nl

0.26

 Netherlands

0.25

van

0.24

Van

0.22

 Amsterdam

0.21

_nl

0.20

recht

0.19

Van

0.19

 amsterdam

0.19

Activations Density 0.243%

references to the Netherlands and Dutch entities

No Comments

No Known Activations