INDEX

Explanations

expressions of gratitude and commendation for experiences or services

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 nice

-0.23

Nice

-0.19

nice

-0.18

æĮº

-0.18

 pleasant

-0.17

 Nice

-0.17

entiful

-0.17

">ÃĹ</

-0.17

pleasant

-0.16

;base

-0.15

POSITIVE LOGITS

 simply

0.39

 beyond

0.31

 Simply

0.30

 incredible

0.29

 amazing

0.29

Simply

0.28

 phenomenal

0.28

 absolutely

0.28

 extraordinary

0.28

 truly

0.28

Activations Density 0.501%

expressions of gratitude and commendation for experiences or services

No Comments

No Known Activations