INDEX

Explanations

the word "under" and its variations, indicating a focus on contexts involving underlying themes or concepts

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

yd

-0.17

wards

-0.16

shaw

-0.16

ify

-0.15

/stretch

-0.15

 ÐºÐ°Ñģ

-0.15

esso

-0.15

ÄĻ

-0.15

urch

-0.15

zw

-0.15

POSITIVE LOGITS

pin

0.29

lie

0.28

 lying

0.26

lying

0.24

lies

0.24

lay

0.23

neath

0.23

lie

0.21

writing

0.19

sea

0.18

Activations Density 0.025%

the word "under" and its variations, indicating a focus on contexts involving underlying themes or concepts

No Comments

No Known Activations