INDEX

Explanations

references to numbers and counting

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

rippling

-0.15

ropoda

-0.14

dbl

-0.13

 Riding

-0.13

kont

-0.13

reste

-0.13

ruh

-0.13

/MIT

-0.13

continuous

-0.13

nÃŃ

-0.13

POSITIVE LOGITS

ago

0.30

 short

0.30

 years

0.26

 scant

0.25

 Christ

0.23

short

0.22

 miles

0.22

 summers

0.22

 months

0.21

 weeks

0.20

Activations Density 0.052%

references to numbers and counting

No Comments

No Known Activations