INDEX

Explanations

segments and parts of a series or structured content

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

ftp

-0.14

åħ°

-0.14

stab

-0.14

 Lane

-0.14

deniz

-0.14

 Heal

-0.14

 casing

-0.14

icken

-0.14

arkin

-0.14

_pid

-0.14

POSITIVE LOGITS

rored

0.17

EDURE

0.16

edor

0.16

uÃ©

0.15

agas

0.14

 PureComponent

0.14

uhe

0.14

omo

0.14

aptop

0.14

ÑĨ

0.14

Activations Density 0.041%

segments and parts of a series or structured content

No Comments

No Known Activations