INDEX

Explanations

references to Swedish cultural or national attributes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

buch

-0.15

reich

-0.15

 voks

-0.14

rech

-0.14

 DÃ¼n

-0.14

 blok

-0.13

iÃ§

-0.13

bish

-0.13

gend

-0.13

 hete

-0.13

POSITIVE LOGITS

 Sweden

0.56

 Swedish

0.54

 Stockholm

0.50

Sweden

0.48

 Svens

0.41

 stockholm

0.38

 Anders

0.35

 Lund

0.34

 Scandinavian

0.32

 Scandin

0.32

Activations Density 0.301%

references to Swedish cultural or national attributes

No Comments

No Known Activations

references to Swedish cultural or national attributes

No Comments

No Known Activations