INDEX

Explanations

mention of the word "simple" and its variations related to concepts and processes

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

 simply

-0.22

 simplement

-0.21

 Simply

-0.20

simple

-0.19

Simply

-0.19

Simple

-0.18

_simple

-0.18

 einfach

-0.18

 simpl

-0.18

 simplified

-0.17

POSITIVE LOGITS

ton

0.43

tons

0.43

xes

0.35

-minded

0.33

TON

0.30

 minded

0.29

ctic

0.27

yet

0.26

/basic

0.24

/plain

0.23

Activations Density 0.040%

mention of the word "simple" and its variations related to concepts and processes

No Comments

No Known Activations