INDEX

Explanations

phrases and structures that involve the word "with" indicating various relationships or emotions

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

 respects

-0.15

 Respect

-0.14

.persistent

-0.14

anych

-0.14

 Î´ÏħÎ½Î±

-0.14

enia

-0.14

ULE

-0.13

/></

-0.13

 Democr

-0.13

herits

-0.13

POSITIVE LOGITS

 abandon

0.33

apl

0.29

 conviction

0.26

 precision

0.26

 gusto

0.25

 ease

0.24

 authority

0.24

 vigor

0.23

 dispatch

0.23

 relative

0.23

Activations Density 0.144%

phrases and structures that involve the word "with" indicating various relationships or emotions

No Comments

No Known Activations