INDEX

Explanations

lists of places, things, or recommendations that are considered the best or top in their category

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

_FD

-0.17

$LANG

-0.17

avourites

-0.16

olars

-0.16

ifestyles

-0.15

ynam

-0.15

 ì§Ī

-0.15

-dismiss

-0.14

edor

-0.14

à¥įà¤¤à¤ķ

-0.14

POSITIVE LOGITS

 places

0.31

 ways

0.26

 Places

0.24

 deals

0.23

 things

0.22

 cheap

0.21

 picks

0.21

 cities

0.21

Places

0.21

 alternatives

0.20

Activations Density 0.101%

lists of places, things, or recommendations that are considered the best or top in their category

No Comments

No Known Activations