INDEX

Explanations

phrases related to questioning or inquiry

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

ood

-0.17

ute

-0.15

RequestMethod

-0.14

Ã¡lido

-0.14

abar

-0.14

(æ°´

-0.14

ainty

-0.14

 trom

-0.14

 Dimit

-0.14

urn

-0.13

POSITIVE LOGITS

 whom

0.27

how

0.23

why

0.22

who

0.22

 ÚĨÚ¯ÙĪÙĨÙĩ

0.20

who

0.20

how

0.19

ä½ķ

0.17

-how

0.17

Who

0.17

Activations Density 0.023%

phrases related to questioning or inquiry

No Comments

No Known Activations