INDEX

Explanations

references to living spaces or areas

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

oples

-0.16

igo

-0.16

 Ð¶Ð¸Ð²

-0.16

ãĥĨãĥ«

-0.15

olini

-0.15

iliar

-0.15

 Lives

-0.15

 Live

-0.15

-lived

-0.14

edly

-0.14

POSITIVE LOGITS

 quarters

0.27

stone

0.21

-wage

0.20

room

0.20

 arrangements

0.20

 wage

0.20

 Wage

0.20

-room

0.19

-quarters

0.19

 expenses

0.18

Activations Density 0.021%

references to living spaces or areas

No Comments

No Known Activations