INDEX

Explanations

inquiries or questions directed at the self or audience

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

opis

-0.17

kup

-0.15

sko

-0.14

ìĿ´ìķ¼

-0.14

kal

-0.13

Ð±ÑĥÐ´ÑĮ

-0.13

ategories

-0.13

loat

-0.13

 Hammond

-0.13

isay

-0.13

POSITIVE LOGITS

0.32

 Does

0.23

 Where

0.23

How

0.23

 What

0.22

Why

0.21

:Is

0.21

Who

0.20

 whether

0.20

 Whether

0.19

Activations Density 0.067%

inquiries or questions directed at the self or audience

No Comments

No Known Activations