INDEX

Explanations

combinations of letters and suffixes commonly associated with scientific names or titles

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 Ø¯Ø§Ø¯

-0.15

/AP

-0.14

anky

-0.14

ÐµÐ´ÑĮ

-0.14

agus

-0.14

isle

-0.13

/U

-0.13

ieten

-0.13

cit

-0.13

zÄĻ

-0.13

POSITIVE LOGITS

ä¸įäºĨ

0.14

¼åĲĪ

0.14

tingham

0.14

_ENSURE

0.13

jamin

0.13

Hi

0.13

DAC

0.13

ures

0.13

ÃĹ</

0.13

à¥įà¤Ĺà¤¤

0.13

Activations Density 0.042%

combinations of letters and suffixes commonly associated with scientific names or titles

No Known Activations