INDEX

Explanations

phrases related to risk and safety 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

fang

-0.15

 Pricing

-0.14

fst

-0.14

uster

-0.14

esian

-0.14

isy

-0.13

aland

-0.13

cak

-0.13

 morphology

-0.13

astes

-0.13

POSITIVE LOGITS

 chances

0.52

 already

0.42

 odds

0.40

already

0.39

 likelihood

0.37

 Already

0.36

 chance

0.35

Already

0.34

likelihood

0.31

_already

0.29

Activations Density 0.165%

phrases related to risk and safety in various contexts

No Comments

No Known Activations