INDEX

Explanations

questions and references to "why" in contexts of confusion or exploration

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.16.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

:numel

-0.15

ule

-0.15

eus

-0.14

.Void

-0.14

ÑĩÐ¸Ñģ

-0.14

ãģ¦ãĤĤ

-0.13

-fontawesome

-0.13

.Bounds

-0.13

agnar

-0.13

etine

-0.13

POSITIVE LOGITS

/how

0.32

 Pant

0.18

soever

0.18

 they

0.16

ulia

0.15

we

0.15

 Mayo

0.14

itzer

0.14

 there

0.14

it

0.14

Activations Density 0.023%

questions and references to "why" in contexts of confusion or exploration

No Comments

No Known Activations