INDEX

Explanations

phrases that express support and guidance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

wy

-0.15

marsh

-0.14

rios

-0.14

ATAL

-0.14

itra

-0.14

ention

-0.14

ä¹ĭä¸Ģ

-0.14

rouw

-0.14

-0.13

alto

-0.13

POSITIVE LOGITS

 through

0.23

 throughout

0.22

 during

0.20

 wherever

0.18

 whenever

0.18

/us

0.17

through

0.16

 closely

0.16

 with

0.16

 THROUGH

0.15

Activations Density 0.155%

phrases that express support and guidance

No Comments

No Known Activations