INDEX

Explanations

references to hosts and hosting situations

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

isto

-0.18

stakes

-0.17

jte

-0.17

pest

-0.16

onest

-0.16

ksam

-0.16

uate

-0.15

jd

-0.15

ãģ°

-0.15

OW

-0.15

POSITIVE LOGITS

ilities

0.43

ess

0.35

etler

0.32

esses

0.31

elry

0.29

ility

0.26

names

0.25

ESS

0.24

els

0.22

essed

0.21

Activations Density 0.029%

references to hosts and hosting situations

No Comments

No Known Activations