INDEX

Explanations

directional and movement-related language

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

ElementException

-0.18

ÅĻÃŃd

-0.16

udd

-0.15

ingly

-0.15

uzey

-0.14

foy

-0.14

ëĬ

-0.14

gi

-0.14

ingo

-0.13

eniable

-0.13

POSITIVE LOGITS

wards

0.24

ward

0.21

 onto

0.21

 into

0.20

WARD

0.17

onto

0.16

oward

0.16

 towards

0.15

 toward

0.15

 Into

0.15

Activations Density 0.158%

directional and movement-related language

No Comments

No Known Activations