INDEX

Explanations

phrases indicating separation or distance from a location or group

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 láº¡i

-0.17

asher

-0.16

pill

-0.16

ovah

-0.16

ual

-0.16

auer

-0.16

lek

-0.15

lady

-0.15

naires

-0.15

ized

-0.15

POSITIVE LOGITS

ward

0.30

wards

0.20

yyyy

0.18

yyy

0.17

ÌĢ

0.16

urved

0.16

oll

0.16

WARD

0.15

gm

0.15

ence

0.15

Activations Density 0.031%

phrases indicating separation or distance from a location or group

No Comments

No Known Activations