INDEX

Explanations

terms related to propaganda and recruitment activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

arget

-0.17

eman

-0.15

errar

-0.15

eken

-0.15

Î²Î¿

-0.14

ernaut

-0.14

aket

-0.14

Î²Î¿Î»

-0.14

 comput

-0.13

ematik

-0.13

POSITIVE LOGITS

incer

0.16

ytic

0.15

_AI

0.15

illis

0.15

Ð¾ÑĢÑĭ

0.14

æĭ©

0.14

ijk

0.14

zcze

0.14

iffs

0.14

yat

0.14

Activations Density 0.036%

terms related to propaganda and recruitment activities

No Comments

No Known Activations