INDEX

Explanations

expressions of gratitude and concern

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

avis

-0.17

ourg

-0.16

ifa

-0.15

ibs

-0.15

ector

-0.15

 ë°Ķë¡ľ

-0.15

abis

-0.14

ufen

-0.14

Ø§Ø¨Ø¹

-0.14

+"<

-0.14

POSITIVE LOGITS

God

0.57

 heaven

0.55

God

0.52

 Heaven

0.50

 heavens

0.50

 goodness

0.50

god

0.49

GOD

0.46

god

0.44

 gods

0.43

Activations Density 0.119%

expressions of gratitude and concern

No Comments

No Known Activations