INDEX

Explanations

references to business-related topics and terms

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

olar

-0.19

pour

-0.16

iper

-0.16

arehouse

-0.16

amed

-0.15

acy

-0.14

atively

-0.14

_nsec

-0.14

VAS

-0.14

orus

-0.14

POSITIVE LOGITS

woman

0.26

people

0.26

 owners

0.24

(es

0.24

person

0.23

women

0.23

persons

0.22

ses

0.21

ness

0.20

 owner

0.20

Activations Density 0.056%

references to business-related topics and terms

No Comments

No Known Activations