INDEX

Explanations

instances of the word "interesting."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

DonaldTrump

-0.15

ldr

-0.15

icaret

-0.15

Ð½Ð³

-0.15

oq

-0.14

lte

-0.14

PTS

-0.14

intosh

-0.13

imas

-0.13

 favorite

-0.13

POSITIVE LOGITS

ekyll

0.17

ly

0.16

emma

0.15

nya

0.15

à¥įà¤µà¤°

0.14

ayi

0.14

WebService

0.14

interesting

0.14

ä¸ľè¥¿

0.14

Activations Density 0.038%

instances of the word "interesting."

No Comments

No Known Activations