INDEX

Explanations

names of people, specifically those with notable initials or last names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

Ð²ÐµÑĤ

-0.16

Interop

-0.16

udad

-0.15

ftware

-0.15

smarty

-0.15

utex

-0.14

urum

-0.14

ices

-0.14

holm

-0.14

stÃŃ

-0.14

POSITIVE LOGITS

 John

0.20

 JOHN

0.17

 Johnny

0.17

John

0.16

 john

0.16

 meta

0.15

ardown

0.15

Joh

0.15

 Seed

0.14

Doe

0.14

Activations Density 0.059%

names of people, specifically those with notable initials or last names

No Comments

No Known Activations