INDEX

Explanations

phrases indicating the highest quality or superiority in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

Ã¡t

-0.16

zap

-0.15

-0.14

 something

-0.14

 nice

-0.14

 matter

-0.14

 favorites

-0.14

ona

-0.14

vt

-0.14

POSITIVE LOGITS

seller

0.34

-selling

0.33

owing

0.32

-known

0.32

 possible

0.29

ow

0.29

-case

0.27

possible

0.26

owment

0.25

ial

0.24

Activations Density 0.049%

phrases indicating the highest quality or superiority in various contexts

No Comments

No Known Activations