INDEX

Explanations

calls to action and promotional phrases

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ekim

-0.18

Ì§

-0.15

OfWork

-0.14

entai

-0.14

lice

-0.13

xAD

-0.13

prak

-0.13

 RSVP

-0.13

Ð¾Ð±ÑĢÐ°Ð

-0.13

ionic

-0.13

POSITIVE LOGITS

 technical

0.26

 Technical

0.26

technical

0.24

 tech

0.23

Technical

0.23

 dial

0.22

 Dial

0.21

 printer

0.20

-tech

0.19

 canon

0.19

Activations Density 0.003%

calls to action and promotional phrases

No Comments

No Known Activations