INDEX

Explanations

content that is pertinent or applicable to a given context

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

rav

-0.19

acters

-0.16

izioni

-0.16

hiba

-0.15

angered

-0.15

asley

-0.15

gb

-0.15

izen

-0.15

erman

-0.15

ocker

-0.15

POSITIVE LOGITS

ly

0.25

äºİ

0.18

/use

0.17

atable

0.17

 äºİ

0.17

/app

0.17

iance

0.17

æĸ¼

0.16

iation

0.16

eting

0.16

Activations Density 0.015%

content that is pertinent or applicable to a given context

No Comments

No Known Activations