INDEX

Explanations

terms related to fences and boundaries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

FileSync

-0.19

ROOM

-0.15

cak

-0.14

éĥ¨å±ĭ

-0.14

_rooms

-0.14

.dashboard

-0.14

itchens

-0.13

olum

-0.13

iler

-0.13

IFO

-0.13

POSITIVE LOGITS

 fencing

0.44

 fence

0.42

 fences

0.39

 Fence

0.36

 gates

0.33

 gate

0.30

 wire

0.29

_fence

0.29

 posts

0.28

 Gates

0.27

Activations Density 0.022%

terms related to fences and boundaries

No Comments

No Known Activations