INDEX

Explanations

references to religious icons and rituals

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

wap

-0.14

grunt

-0.14

%"><

-0.13

LETTE

-0.13

 clearColor

-0.13

ennie

-0.13

ì¶Ķ

-0.13

qui

-0.13

Å¼y

-0.12

 demean

-0.12

POSITIVE LOGITS

 capitals

0.23

 bronze

0.22

 gold

0.22

 overl

0.22

 cher

0.21

 overlay

0.21

 hammered

0.20

 cedar

0.20

 ornament

0.20

cub

0.19

Activations Density 0.030%

references to religious icons and rituals

No Comments

No Known Activations