INDEX

Explanations

URLs and links to stories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

_KP

-0.17

hur

-0.15

-0.14

INI

-0.14

ials

-0.14

ika

-0.14

inta

-0.14

rad

-0.14

ini

-0.14

 wears

-0.14

POSITIVE LOGITS

Helmet

0.15

ugen

0.15

 sanction

0.15

efe

0.15

arshal

0.15

ycz

0.14

efs

0.14

izin

0.14

olley

0.14

ì£¼

0.14

Activations Density 0.001%

URLs and links to stories

No Comments

No Known Activations