INDEX

Explanations

terms related to increments or enhancements 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 @ 25-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.25.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

ALSE

-0.15

ersh

-0.15

elter

-0.15

 Unary

-0.14

ynamics

-0.14

à¥Įà¤¨

-0.14

olicies

-0.14

VERR

-0.14

 Timing

-0.14

 demographics

-0.13

POSITIVE LOGITS

 presence

0.28

 emphasis

0.28

 attention

0.25

 focus

0.25

 awareness

0.23

presence

0.23

 involvement

0.22

emphasis

0.22

 participation

0.21

 investment

0.21

Activations Density 0.206%

terms related to increments or enhancements in various contexts

No Comments

No Known Activations