INDEX

Explanations

phrases that refer to actions users can take, such as following, reading, or listening

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

 inst

-0.16

iro

-0.15

anny

-0.15

rob

-0.15

ventus

-0.15

 favor

-0.15

 ifndef

-0.14

obra

-0.14

 favour

-0.14

annels

-0.14

POSITIVE LOGITS

lauf

0.17

 Voll

0.16

ovnÃŃ

0.16

ctors

0.15

lander

0.14

istrovstvÃŃ

0.14

yonel

0.14

oproject

0.14

.Err

0.14

.Magenta

0.14

Activations Density 0.058%

phrases that refer to actions users can take, such as following, reading, or listening

No Comments

No Known Activations