INDEX

Explanations

the word "empower" and its variants, indicating a focus on empowerment themes

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

inks

-0.16

pst

-0.16

ration

-0.15

 extern

-0.15

bum

-0.15

bsp

-0.14

uba

-0.14

inx

-0.14

erne

-0.14

eward

-0.14

POSITIVE LOGITS

irical

0.34

owered

0.33

owering

0.32

yre

0.26

ower

0.26

LOYEE

0.25

athy

0.25

LOY

0.24

oleon

0.22

orio

0.22

Activations Density 0.010%

the word "empower" and its variants, indicating a focus on empowerment themes

No Comments

No Known Activations