INDEX

Explanations

references to brands and companies

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

iec

-0.17

idge

-0.17

imar

-0.16

uby

-0.16

Pros

-0.15

 studio

-0.15

etto

-0.14

ije

-0.13

ridge

-0.13

_SI

-0.13

POSITIVE LOGITS

 Steel

0.20

Steel

0.19

 Tata

0.18

steel

0.17

 steel

0.17

ons

0.16

awks

0.16

zp

0.16

 salt

0.16

 passenger

0.16

Activations Density 0.003%

references to brands and companies

No Comments

No Known Activations