INDEX

Explanations

historical references related to architecture and war events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

 à¤Ńà¤Ĺ

-0.16

 Pazar

-0.16

PAT

-0.15

iene

-0.15

 Libyan

-0.15

UCT

-0.14

PUR

-0.14

iores

-0.14

POSITIVE LOGITS

 Louis

0.22

 Cardinal

0.20

Lau

0.17

 Font

0.17

 Baum

0.16

 Bourbon

0.15

 Loud

0.15

 campaigning

0.15

Da

0.15

 camp

0.15

Activations Density 0.008%

historical references related to architecture and war events

No Comments

No Known Activations