INDEX

Explanations

numerical values and specific dates or ordinal indicators

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 favors

-0.17

 Favorite

-0.17

 colorful

-0.16

-percent

-0.16

eto

-0.16

 honor

-0.16

Mom

-0.16

ursed

-0.15

 honors

-0.15

 Pron

-0.15

POSITIVE LOGITS

|_|

0.16

Formats

0.14

 Guild

0.14

ben

0.14

uro

0.13

 Jako

0.13

ãģŃ

0.13

ÙĨÙĩ

0.13

éĥ

0.13

é£

0.13

Activations Density 0.120%

numerical values and specific dates or ordinal indicators

No Comments

No Known Activations