INDEX

Explanations

phrases that indicate essential recommendations or must-have items

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

ninger

-0.17

annes

-0.16

olson

-0.16

lets

-0.15

istrovstvÃŃ

-0.15

dez

-0.15

Ð±Ð¾

-0.15

 Lever

-0.15

Yug

-0.15

illin

-0.14

POSITIVE LOGITS

ache

0.28

-have

0.28

ering

0.25

afa

0.23

ached

0.22

aches

0.22

ACHE

0.21

angs

0.21

aph

0.19

Activations Density 0.014%

phrases that indicate essential recommendations or must-have items

No Comments

No Known Activations