INDEX

Explanations

phrases expressing approval or positive evaluations

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

dit

-0.18

yonel

-0.18

yms

-0.17

ein

-0.17

elli

-0.17

elect

-0.17

yen

-0.16

ModelError

-0.16

eum

-0.16

yll

-0.16

POSITIVE LOGITS

-known

0.33

spring

0.31

ington

0.29

-being

0.27

come

0.26

ows

0.25

 enough

0.24

-rounded

0.23

llll

0.23

known

0.23

Activations Density 0.067%

phrases expressing approval or positive evaluations

No Comments

No Known Activations