INDEX

Explanations

phrases that emphasize similarity and shared experiences

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

ocker

-0.16

erland

-0.15

\Php

-0.14

ington

-0.14

eric

-0.14

ialog

-0.14

mony

-0.14

erus

-0.14

uckles

-0.14

ÃŁer

-0.14

POSITIVE LOGITS

 identical

0.43

 similar

0.40

çĽ¸åĲĮ

0.38

 alike

0.35

åĲĮãģĺ

0.34

same

0.34

 same

0.34

åĲĮ

0.33

 Same

0.33

similar

0.32

Activations Density 0.223%

phrases that emphasize similarity and shared experiences

No Comments

No Known Activations