INDEX

Explanations

Boolean values indicating truth or existence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

lar

-0.16

prot

-0.15

locator

-0.15

ship

-0.15

stÃŃ

-0.14

-upper

-0.14

.gg

-0.14

 Upper

-0.14

 nackte

-0.14

.elementAt

-0.14

POSITIVE LOGITS

/false

0.27

/null

0.18

/*

0.16

.obs

0.15

zos

0.15

 Keller

0.15

/*!

0.15

/*!

0.15

Activations Density 0.103%

Boolean values indicating truth or existence

No Comments

No Known Activations