INDEX

Explanations

references to popular cartoon characters and related themes

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

phet

-0.16

ossa

-0.15

/import

-0.15

ovsky

-0.14

.Middle

-0.14

Ã¡no

-0.13

vero

-0.13

Ø®ÙĦ

-0.13

ì¶ķ

-0.13

:convert

-0.13

POSITIVE LOGITS

 Naruto

0.24

å¿į

0.22

nin

0.20

Nin

0.20

 ninja

0.19

nin

0.16

 Sense

0.16

 Ninja

0.16

 dojo

0.16

zen

0.15

Activations Density 0.027%

references to popular cartoon characters and related themes

No Comments

No Known Activations