INDEX

Explanations

instances of the phrase "So" with varying degrees of frequency or prominence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 17-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L17R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.17.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

so

-0.18

quam

-0.17

ugar

-0.17

esome

-0.16

ifestyles

-0.15

 ÑĤÐ°Ðº

-0.15

coon

-0.15

èŀį

-0.15

pane

-0.15

ãĥ³ãĥĲãĥ¼

-0.15

POSITIVE LOGITS

-called

0.32

ftware

0.27

ething

0.23

far

0.23

 much

0.21

 sÃ¡nh

0.20

cial

0.20

theast

0.20

ilder

0.19

 Much

0.18

Activations Density 0.044%

instances of the phrase "So" with varying degrees of frequency or prominence

No Comments

No Known Activations