INDEX

Explanations

references to military rank and service history

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

idges

-0.17

dorf

-0.15

roker

-0.14

anke

-0.14

acks

-0.14

 pitch

-0.14

kar

-0.14

anner

-0.14

_crit

-0.14

airport

-0.14

POSITIVE LOGITS

oler

0.15

gratis

0.14

 Rapids

0.14

grily

0.14

rape

0.14

 alma

0.14

ikit

0.14

 itemprop

0.13

 Sloan

0.13

 Rape

0.13

Activations Density 0.009%

references to military rank and service history

No Comments

No Known Activations