INDEX

Explanations

numerical identifiers and addresses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

ight

-0.16

 Walton

-0.16

Ã¤ng

-0.15

.scalablytyped

-0.15

ieties

-0.15

olia

-0.15

ients

-0.14

_PD

-0.14

iedad

-0.14

Dew

-0.14

POSITIVE LOGITS

uida

0.16

akit

0.15

 dome

0.14

onse

0.14

enna

0.14

Ð°Ð¼Ð¿

0.14

uns

0.14

odom

0.14

 nale

0.14

lander

0.14

Activations Density 0.070%

numerical identifiers and addresses

No Comments

No Known Activations