INDEX

Explanations

occurrences of the name "Le" or similar variants, indicating a focus on specific individuals or surnames

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

gli

-0.17

zej

-0.17

gene

-0.15

Ã©ra

-0.15

rix

-0.15

iffin

-0.15

rtl

-0.14

CoreApplication

-0.14

ncy

-0.14

POSITIVE LOGITS

Roy

0.24

athers

0.24

Ã³n

0.23

Page

0.22

ighton

0.22

andro

0.22

wis

0.22

Rou

0.21

Roy

0.21

hots

0.21

Activations Density 0.017%

occurrences of the name "Le" or similar variants, indicating a focus on specific individuals or surnames

No Comments

No Known Activations