INDEX

Explanations

references to political parties and governmental discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

HAL

-0.15

County

-0.15

lius

-0.15

 filmy

-0.14

 Borough

-0.14

éĥ¡

-0.14

THON

-0.14

 Raqqa

-0.14

 County

-0.14

Ã¼yÃ¼k

-0.14

POSITIVE LOGITS

 Pres

0.23

 Holy

0.22

 Conv

0.20

 conv

0.19

Cos

0.18

Conv

0.18

Hum

0.18

COS

0.18

 First

0.18

 Finance

0.18

Activations Density 0.003%

references to political parties and governmental discussions

No Comments

No Known Activations