INDEX

Explanations

programming-related terminology, particularly pertaining to object-oriented design and data management

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

abus

-0.17

orum

-0.16

MT

-0.15

obble

-0.14

Geb

-0.14

anity

-0.14

Vtbl

-0.14

lingen

-0.14

avor

-0.14

Ã©ric

-0.14

POSITIVE LOGITS

wx

0.20

(wx

0.20

=wx

0.19

(IDC

0.17

IMPLEMENT

0.16

 wxDefault

0.16

ProgressDialog

0.16

wx

0.15

AFX

0.15

wel

0.15

Activations Density 0.004%

programming-related terminology, particularly pertaining to object-oriented design and data management

No Comments

No Known Activations