INDEX

Explanations

content related to performing arts training programs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 27-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L27R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.27.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

 ÑģÑħÐ¾Ð´

-0.15

 fisse

-0.15

vido

-0.15

 Abyss

-0.15

éĽĦ

-0.14

 PROCUREMENT

-0.14

orb

-0.13

 engel

-0.13

nh

-0.13

 Terrace

-0.13

POSITIVE LOGITS

 Impro

0.25

 improv

0.23

Impro

0.22

 Acting

0.21

 improvis

0.20

 Stan

0.20

 impro

0.20

 improvised

0.20

 Blocking

0.20

 spont

0.20

Activations Density 0.040%

content related to performing arts training programs

No Comments

No Known Activations