INDEX

Explanations

references to door-related products or features

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

uxt

-0.17

ulo

-0.16

 hier

-0.15

altar

-0.15

/proto

-0.15

ÑĤÑĮ

-0.14

ussian

-0.14

gency

-0.14

orsi

-0.14

ëĿ¼ìĿ´

-0.14

POSITIVE LOGITS

 door

0.19

 Door

0.18

-door

0.17

Door

0.17

 hinges

0.16

 installed

0.16

 installation

0.16

0.15

 opening

0.15

 opener

0.15

Activations Density 0.037%

references to door-related products or features

No Comments

No Known Activations