INDEX

Explanations

various forms of the word "post"

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

vÃŃ

-0.18

rats

-0.15

apa

-0.15

rat

-0.15

bsolute

-0.15

zelf

-0.15

olds

-0.14

vier

-0.14

entifier

-0.14

lesh

-0.14

POSITIVE LOGITS

ulate

0.38

mort

0.32

graduate

0.31

greSQL

0.31

ulates

0.31

natal

0.30

ulated

0.30

cards

0.29

uring

0.28

pone

0.28

Activations Density 0.069%

various forms of the word "post"

No Comments

No Known Activations