INDEX

Explanations

details about authorship and posting information in a blog format

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ÐµÐºÑĤÐ¾ÑĢ

-0.15

ons

-0.15

onn

-0.15

Ø±ÛĮÙħ

-0.14

prix

-0.14

visor

-0.14

chodu

-0.14

Ã§on

-0.14

toJson

-0.14

adesh

-0.14

POSITIVE LOGITS

 Leave

0.21

Leave

0.21

 leave

0.19

obil

0.17

leave

0.17

apat

0.15

 leaves

0.15

ein

0.15

.leave

0.15

oub

0.15

Activations Density 0.016%

details about authorship and posting information in a blog format

No Comments

No Known Activations