INDEX

Explanations

numbers and dates associated with events or historical records

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

isman

-0.15

actics

-0.14

 reception

-0.14

 expos

-0.14

anki

-0.14

Obt

-0.14

ropolis

-0.14

 thá»ĭ

-0.13

 scratch

-0.13

iniz

-0.13

POSITIVE LOGITS

 gene

0.22

 Chronicles

0.20

 Histor

0.20

 Camden

0.19

 History

0.19

 Rolls

0.19

 Monument

0.19

Roll

0.18

 Chron

0.18

 chronic

0.18

Activations Density 0.086%

numbers and dates associated with events or historical records

No Comments

No Known Activations