INDEX

Explanations

words and phrases related to compliance and regulations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gles

-0.19

 adher

-0.18

COM

-0.17

COM

-0.16

 treff

-0.16

ãĤ³ãĥ³

-0.15

com

-0.15

(disposing

-0.15

iben

-0.15

.componentInstance

-0.14

POSITIVE LOGITS

 conform

0.34

pliant

0.33

 Compliance

0.33

Ad

0.32

ad

0.30

 obedience

0.30

 compliance

0.30

pliance

0.29

 conformity

0.28

formance

0.28

Activations Density 0.005%

words and phrases related to compliance and regulations

No Comments

No Known Activations