INDEX

Explanations

references to deception or dishonesty

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

Ð¾ÑĢÐ¾Ð²

-0.16

-blind

-0.16

sar

-0.16

íģ¼

-0.16

.scalablytyped

-0.15

tron

-0.15

ousse

-0.14

 blind

-0.14

meer

-0.14

blind

-0.14

POSITIVE LOGITS

uten

0.24

utenant

0.20

 about

0.18

_about

0.17

 detector

0.17

-flat

0.17

chten

0.17

urance

0.16

Lie

0.16

 Detector

0.16

Activations Density 0.016%

references to deception or dishonesty

No Comments

No Known Activations