INDEX

Explanations

phrases and concepts related to sizes and fitting

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

 Folk

-0.20

 Feder

-0.20

Fallback

-0.20

 Fellowship

-0.19

 Favor

-0.19

 Folder

-0.18

Fol

-0.17

 Friendship

-0.17

Fur

-0.17

 Flooring

-0.17

POSITIVE LOGITS

fit

0.73

-fit

0.63

-f

0.57

 fits

0.57

fit

0.57

.fit

0.52

Fit

0.51

_fit

0.51

Fit

0.50

FIT

0.45

Activations Density 0.122%

phrases and concepts related to sizes and fitting

No Comments

No Known Activations