INDEX

Explanations

references to possession and relationships between characters and elements in the text

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 byÅĤo

-0.18

uji

-0.15

 ballo

-0.15

 generado

-0.15

 completo

-0.14

 giorno

-0.14

bij

-0.14

 nuovo

-0.14

chio

-0.14

 dolore

-0.14

POSITIVE LOGITS

 misma

0.23

zos

0.21

itud

0.17

 dÃ©c

0.17

zo

0.17

icode

0.16

ughs

0.16

uds

0.15

zar

0.15

 Ã©p

0.15

Activations Density 0.032%

references to possession and relationships between characters and elements in the text

No Comments

No Known Activations