INDEX

Explanations

phrases that refer to annual events or conferences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

pez

-0.17

elf

-0.14

ov

-0.14

etz

-0.14

 second

-0.14

 segunda

-0.13

ihn

-0.13

upp

-0.13

AFE

-0.13

ort

-0.13

POSITIVE LOGITS

 annual

0.33

 edition

0.32

 Annual

0.30

annual

0.28

Annual

0.27

 iteration

0.25

å±Ĩ

0.25

 editions

0.24

edition

0.23

Edition

0.21

Activations Density 0.043%

phrases that refer to annual events or conferences

No Comments

No Known Activations