INDEX

Explanations

instances of quotations or dialogue in a 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

 unless

-0.16

if

-0.16

moz

-0.15

holm

-0.15

unless

-0.15

 then

-0.15

kiye

-0.14

Can

-0.14

instead

-0.14

uiten

-0.14

POSITIVE LOGITS

 alike

0.21

having

0.19

 having

0.19

 Having

0.17

 Partner

0.17

 hearing

0.17

Congratulations

0.17

Having

0.17

partner

0.17

Partner

0.16

Activations Density 0.073%

instances of quotations or dialogue in a text

No Comments

No Known Activations