INDEX

Explanations

geographic references and terms related to military actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

.mx

-0.21

PÅĻÃŃ

-0.19

 KrÃ¡l

-0.16

PÅĻed

-0.15

Ð»Ð¸Ð²

-0.15

Streamer

-0.15

/epl

-0.15

bjerg

-0.15

otton

-0.15

ROID

-0.15

POSITIVE LOGITS

 Ukraine

0.34

 Ukrainian

0.32

 Ukrain

0.28

Ukr

0.28

 Kiev

0.26

Lv

0.25

 Baltimore

0.23

 Donetsk

0.22

 Crimea

0.22

 Ukraj

0.22

Activations Density 0.074%

geographic references and terms related to military actions

No Comments

No Known Activations