INDEX

Explanations

references to cults or cult-like behavior

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

ctype

-0.16

lijke

-0.16

ansen

-0.16

pector

-0.15

pii

-0.15

ieties

-0.15

ITY

-0.15

annes

-0.14

ç«ĭãģ¦

-0.14

/octet

-0.14

POSITIVE LOGITS

urally

0.35

ivating

0.34

ivate

0.33

ivated

0.32

ivation

0.31

ured

0.28

URAL

0.27

ish

0.26

ists

0.25

ivar

0.25

Activations Density 0.007%

references to cults or cult-like behavior

No Comments

No Known Activations