INDEX

Explanations

technical terminology related to web development processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 6-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L6R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.6.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

 ÎĶÎµ

-0.17

riere

-0.15

owski

-0.15

-0.14

bal

-0.14

.maxLength

-0.14

íķĢ

-0.14

iglia

-0.14

.rollback

-0.14

è¼

-0.14

POSITIVE LOGITS

egin

0.16

 Entry

0.15

Entry

0.15

ÃŃd

0.15

.nl

0.15

æ´¾

0.15

odge

0.15

agate

0.14

ĨĴ

0.14

exit

0.14

Activations Density 0.158%

technical terminology related to web development processes

No Comments

No Known Activations