INDEX

Explanations

instances of the word "build" and its variations, indicating a focus on construction or creation activities

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

er

-0.23

.idea

-0.17

none

-0.16

es

-0.16

iras

-0.16

erer

-0.15

ocha

-0.15

dÃ¼

-0.15

zon

-0.15

als

-0.15

POSITIVE LOGITS

 upon

0.30

Upon

0.24

 dá»±ng

0.23

 Upon

0.23

-up

0.22

.VERSION

0.21

 Blocks

0.21

-in

0.20

ings

0.20

-block

0.19

Activations Density 0.066%

instances of the word "build" and its variations, indicating a focus on construction or creation activities

No Comments

No Known Activations