INDEX

Explanations

words related to threats and risks, particularly in the context of safety and health

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

uster

-0.17

etch

-0.17

ETCH

-0.17

agan

-0.16

rends

-0.15

uzey

-0.15

 Mahar

-0.14

amar

-0.14

.UnitTesting

-0.14

um

-0.14

POSITIVE LOGITS

 danger

0.20

 dangers

0.20

 Danger

0.18

 Threat

0.16

itzer

0.16

ascus

0.15

geg

0.15

Ŀ

0.15

Danger

0.15

 risks

0.15

Activations Density 0.150%

words related to threats and risks, particularly in the context of safety and health

No Comments

No Known Activations