INDEX

Explanations

expressions of willingness and helpfulness towards communication and assistance

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

iedo

-0.17

allah

-0.17

áº¡i

-0.15

nist

-0.15

ierz

-0.14

oldem

-0.14

allo

-0.14

Ø§Ø¦ÛĮ

-0.14

gunakan

-0.14

IRM

-0.13

POSITIVE LOGITS

 happy

0.41

 Happy

0.36

Happy

0.35

happy

0.35

 happiness

0.34

 HAPP

0.31

 glad

0.30

 Happiness

0.28

 Glad

0.27

 pleasure

0.26

Activations Density 0.052%

expressions of willingness and helpfulness towards communication and assistance

No Comments

No Known Activations