INDEX

Explanations

occurrences of the word "one" and its variations

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

letal

-0.15

omite

-0.15

EventManager

-0.14

aires

-0.14

oku

-0.14

eps

-0.14

/wiki

-0.14

eus

-0.14

ural

-0.14

å¤§åħ¨

-0.14

POSITIVE LOGITS

 reason

0.27

 might

0.26

 thing

0.24

 cannot

0.23

of

0.22

can

0.21

 could

0.21

 such

0.20

 advantage

0.19

 need

0.19

Activations Density 0.054%

occurrences of the word "one" and its variations

No Comments

No Known Activations