INDEX

Explanations

instances of the word "let" and its variations, as well as the word "much."

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

 Allow

-0.25

_allow

-0.22

 lets

-0.21

 Lets

-0.21

Allow

-0.21

 letting

-0.21

lets

-0.20

 allow

-0.20

 ALLOW

-0.20

let

-0.19

POSITIVE LOGITS

 alone

0.33

 least

0.26

 less

0.24

alone

0.23

Least

0.22

 Least

0.21

 Alone

0.21

less

0.20

 menos

0.20

least

0.19

Activations Density 0.008%

instances of the word "let" and its variations, as well as the word "much."

No Comments

No Known Activations