INDEX

Explanations

references to storage locations and household items

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

åºŃ

-0.16

istik

-0.15

 Chain

-0.14

åº§

-0.14

Ã¥de

-0.14

 Chairs

-0.14

ibi

-0.14

Forum

-0.13

platz

-0.13

 Chains

-0.13

POSITIVE LOGITS

 storage

0.37

 drawer

0.33

 trunk

0.33

 closet

0.30

 Storage

0.30

 drawers

0.30

 boxes

0.30

 cupboard

0.29

storage

0.28

 cabinet

0.28

Activations Density 0.193%

references to storage locations and household items

No Comments

No Known Activations