INDEX

Explanations

phrases about locations or contexts, particularly in relation to temperature and environmental settings

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 light

-0.20

 Light

-0.16

Ã¦re

-0.16

junction

-0.15

akening

-0.15

 Lights

-0.14

 LIGHT

-0.14

uing

-0.14

caret

-0.14

 lights

-0.14

POSITIVE LOGITS

bad

0.20

lay

0.20

lets

0.19

board

0.18

inc

0.17

 poor

0.17

 warmer

0.17

 foul

0.17

 warm

0.17

 situ

0.17

Activations Density 0.295%

phrases about locations or contexts, particularly in relation to temperature and environmental settings

No Known Activations