INDEX

Explanations

concepts related to meta and philosophical discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

_marshall

-0.16

ÐµÐ»ÐµÑĦ

-0.15

edException

-0.15

ï¿¥

-0.15

rowse

-0.15

ãĥĶãĥ¼

-0.15

(æ°´

-0.15

opensource

-0.14

addon

-0.14

çĽĳåĲ¬é¡µéĿ¢

-0.14

POSITIVE LOGITS

 anti

0.33

 pseudo

0.31

 quasi

0.30

 Anti

0.29

 semi

0.29

anti

0.28

Anti

0.27

 Semi

0.26

pseudo

0.23

 proto

0.23

Activations Density 0.153%

concepts related to meta and philosophical discussions

No Comments

No Known Activations