INDEX

Explanations

relationships involving contrasting or comparative elements

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

andest

-0.18

Ð¸ÑĤÐ¾ÑĢ

-0.15

 dziewcz

-0.14

 Ð¿ÐµÑĢÐµÐ»

-0.14

uyá»ĥn

-0.14

/renderer

-0.14

steder

-0.13

 Stateless

-0.13

 Unsigned

-0.13

lient

-0.13

POSITIVE LOGITS

 veteran

0.66

 veterans

0.59

 seasoned

0.56

 experienced

0.56

 veter

0.53

 Veteran

0.49

Veter

0.48

 Experienced

0.42

vet

0.41

 Veterans

0.40

Activations Density 0.195%

relationships involving contrasting or comparative elements

No Comments

No Known Activations