INDEX

Explanations

mentions of law enforcement and legal terminology

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

arium

-0.18

Ã¤t

-0.17

iban

-0.16

uggling

-0.16

ilot

-0.15

uja

-0.15

acting

-0.15

Traits

-0.15

polit

-0.15

opian

-0.15

POSITIVE LOGITS

 enforcement

0.46

 Enforcement

0.39

fully

0.27

 enforce

0.27

 enforcing

0.27

fulness

0.25

rence

0.25

forcement

0.23

-ab

0.23

 enforced

0.22

Activations Density 0.010%

mentions of law enforcement and legal terminology

No Comments

No Known Activations