INDEX

Explanations

representations of various header levels or titles in text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

eh

-0.17

hips

-0.16

iment

-0.14

imentary

-0.14

faction

-0.14

mÃ¼r

-0.14

ellar

-0.14

hard

-0.14

-0.13

aub

-0.13

POSITIVE LOGITS

soever

0.15

OwnProperty

0.15

ROL

0.15

 Brace

0.14

ry

0.14

.CustomButton

0.14

keit

0.14

InParameter

0.14

.mybatisplus

0.13

ork

0.13

Activations Density 0.128%

representations of various header levels or titles in text

No Comments

No Known Activations