INDEX

Explanations

references to formal structures or institutions

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

-than

-0.15

iÐ´

-0.15

aviest

-0.14

ccess

-0.14

ght

-0.13

_listener

-0.13

iable

-0.13

thur

-0.13

izable

-0.13

ÑĥÑī

-0.13

POSITIVE LOGITS

etheless

0.19

Ã¢ce

0.18

lieÃŁlich

0.17

puÃ©s

0.17

ÃŁerdem

0.16

 Kostenlose

0.16

 wiÄĻc

0.15

orem

0.15

kea

0.15

sequently

0.14

Activations Density 0.107%

references to formal structures or institutions

No Comments

No Known Activations