INDEX

Explanations

terms related to film, architecture, and education

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

iaux

-0.20

 Spears

-0.15

uguay

-0.15

 nues

-0.15

-0.14

erken

-0.14

iversit

-0.14

adol

-0.13

 à¤ħà¤ĸ

-0.13

otta

-0.13

POSITIVE LOGITS

 stub

0.17

 Hanna

0.17

 Â§

0.16

ouchers

0.14

antan

0.14

atern

0.14

DOMNode

0.13

çļ®

0.13

IDA

0.13

ter

0.13

Activations Density 0.019%

terms related to film, architecture, and education

No Comments

No Known Activations