INDEX

Explanations

phrases related to admission or acknowledgement of wrongdoing

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L1R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.1.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

xia

-0.17

 Levine

-0.17

/UIKit

-0.16

ahu

-0.16

 Lump

-0.15

orro

-0.15

(PR

-0.15

 Bilim

-0.14

arya

-0.14

 incon

-0.14

POSITIVE LOGITS

mlin

0.17

0.14

ietet

0.14

itto

0.14

dvd

0.13

forced

0.13

 Hayward

0.13

istrovstvÃŃ

0.13

TestCategory

0.13

iets

0.13

Activations Density 0.002%

phrases related to admission or acknowledgement of wrongdoing

No Comments

No Known Activations