INDEX

Explanations

references to the name "Whit" in various contexts

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ertz

-0.18

ive

-0.17

es

-0.16

signature

-0.16

sWith

-0.16

al

-0.15

ion

-0.15

ei

-0.15

atinum

-0.15

POSITIVE LOGITS

aker

0.29

ewater

0.27

akers

0.24

marsh

0.23

AKER

0.23

ney

0.23

bread

0.23

etail

0.22

acre

0.21

ening

0.21

Activations Density 0.005%

references to the name "Whit" in various contexts

No Known Activations