INDEX

Explanations

occurrences of specific numerical or citation structures within the text

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

 æģ

-0.17

achs

-0.15

gal

-0.15

reau

-0.15

tring

-0.15

 Sight

-0.15

rieb

-0.14

 sight

-0.14

Surname

-0.14

 ÑıÑģ

-0.14

POSITIVE LOGITS

van

0.17

Alo

0.17

 Gent

0.15

tre

0.15

Tre

0.15

Ã¶st

0.14

 Ment

0.14

âĢ¦↵↵↵

0.14

oce

0.14

Ã¡vka

0.14

Activations Density 0.000%

occurrences of specific numerical or citation structures within the text

No Comments

No Known Activations