INDEX

Explanations

references to discussions or panels related to social or historical topics

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

iciel

-0.15

alg

-0.15

 conventions

-0.15

anova

-0.14

 contained

-0.14

 Christmas

-0.14

 christmas

-0.13

á»ī

-0.13

epy

-0.13

 deleg

-0.13

POSITIVE LOGITS

 panel

0.27

 conversation

0.25

panel

0.25

 discussion

0.24

 moderation

0.23

 moderated

0.22

 moder

0.22

forum

0.22

conversation

0.22

 Moder

0.21

Activations Density 0.106%

references to discussions or panels related to social or historical topics

No Comments

No Known Activations