INDEX

Explanations

references to specific mystery book titles and their associated characters

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

ç®

-0.15

 filib

-0.15

lore

-0.14

ENCHMARK

-0.14

ÃŁer

-0.14

ema

-0.14

 ØªÙĤÙĪ

-0.14

hma

-0.13

 Define

-0.13

endir

-0.13

POSITIVE LOGITS

 crime

0.41

 Crime

0.34

 detective

0.33

Crime

0.32

crime

0.31

 detectives

0.31

 mystery

0.30

 mysteries

0.30

 crimes

0.28

 police

0.28

Activations Density 0.181%

references to specific mystery book titles and their associated characters

No Comments

No Known Activations