INDEX

Explanations

requests for help and advice

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

.toFloat

-0.15

Éµ

-0.15

oir

-0.15

inspace

-0.15

inst

-0.15

AsyncResult

-0.14

illet

-0.14

KANJI

-0.14

_DIP

-0.14

iel

-0.14

POSITIVE LOGITS

 help

0.37

 suggestions

0.35

 assistance

0.31

 suggestion

0.30

help

0.30

 ideas

0.30

 Help

0.29

-help

0.27

 advice

0.27

.help

0.27

Activations Density 0.041%

requests for help and advice

No Comments

No Known Activations