INDEX

Explanations

phrases related to structural and functional components

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

eken

-0.15

ecd

-0.14

imag

-0.14

-ton

-0.14

omaly

-0.14

Î¹Î½ÏĮ

-0.14

elle

-0.14

 light

-0.14

trak

-0.14

-large

-0.14

POSITIVE LOGITS

lay

0.22

 situ

0.22

 sizes

0.15

osate

0.15

ibri

0.15

 place

0.15

idl

0.15

ANNER

0.15

erties

0.14

ç½

0.14

Activations Density 0.254%

phrases related to structural and functional components

No Comments

No Known Activations