INDEX

Explanations

references to historical events and structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

Mountain

-0.17

+xml

-0.16

achs

-0.15

 Turnbull

-0.15

atism

-0.14

 Mountain

-0.14

ibName

-0.14

 McMaster

-0.14

 åĩ

-0.14

-www

-0.13

POSITIVE LOGITS

 English

0.20

 England

0.20

 Henry

0.19

 Charter

0.19

 Wyatt

0.18

 english

0.18

 London

0.18

England

0.18

English

0.17

 Surrey

0.17

Activations Density 0.146%

references to historical events and structures

No Comments

No Known Activations