INDEX

Explanations

words associated with electricity and its related concepts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.23

ing

-0.21

ings

-0.20

à¸£

-0.19

scape

-0.19

tings

-0.19

hook

-0.19

suit

-0.18

Ùĩ

-0.18

iest

-0.18

POSITIVE LOGITS

ALLY

0.64

ally

0.57

ity

0.40

amente

0.32

all

0.32

ITY

0.31

ians

0.28

ated

0.27

ities

0.26

ating

0.25

Activations Density 0.200%

words associated with electricity and its related concepts

No Comments

No Known Activations