INDEX

Explanations

historical references and connections to specific events or figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

aab

-0.14

ÎµÏģ

-0.14

 Analog

-0.14

 forth

-0.13

 OPTIONAL

-0.13

rets

-0.13

 Stuff

-0.13

rika

-0.13

eto

-0.13

ponde

-0.13

POSITIVE LOGITS

 records

0.42

 record

0.37

records

0.36

record

0.35

 Records

0.35

Records

0.33

 recorded

0.32

 RECORD

0.31

-record

0.30

 Record

0.29

Activations Density 0.301%

historical references and connections to specific events or figures

No Comments

No Known Activations