INDEX

Explanations

phrases related to subscriptions and notifications

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

ogo

-0.18

ERSION

-0.15

 sublicense

-0.14

 Äĳáº¡i

-0.14

ourn

-0.14

ime

-0.14

nown

-0.14

quals

-0.14

iances

-0.13

zak

-0.13

POSITIVE LOGITS

 exclusive

0.27

 updates

0.26

-exclusive

0.22

exclusive

0.22

 future

0.21

 occasional

0.21

 alerts

0.21

 regular

0.20

 notifications

0.20

 breaking

0.20

Activations Density 0.060%

phrases related to subscriptions and notifications

No Comments

No Known Activations