INDEX

Explanations

variations of the word "gig" or related terms indicating size or scale

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 23-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L23R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.23.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

Ø©

-0.18

iffe

-0.17

ãģ£ãģį

-0.16

iffer

-0.16

ras

-0.16

nell

-0.16

chief

-0.16

asive

-0.15

aab

-0.15

POSITIVE LOGITS

gle

0.27

ging

0.27

gy

0.25

keiten

0.24

gin

0.22

ged

0.19

ean

0.19

gest

0.19

keit

0.18

antically

0.18

Activations Density 0.046%

variations of the word "gig" or related terms indicating size or scale

No Comments

No Known Activations