INDEX

Explanations

references to the concept of power and empowerment

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

_power

-0.19

 power

-0.18

sis

-0.18

POWER

-0.18

 poder

-0.17

 POWER

-0.17

 pouvoir

-0.17

suppress

-0.16

abei

-0.16

aryawan

-0.16

POSITIVE LOGITS

fully

0.42

houses

0.31

ful

0.29

full

0.28

train

0.25

lifting

0.24

FUL

0.23

lessness

0.22

plant

0.21

ment

0.20

Activations Density 0.063%

references to the concept of power and empowerment

No Comments

No Known Activations