INDEX

Explanations

references to web links and numeric data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 13-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L13R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.13.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

(CultureInfo

-0.16

-0.14

lett

-0.14

otron

-0.14

ç«¹

-0.14

iÄįka

-0.14

gh

-0.14

ucch

-0.14

arked

-0.14

abe

-0.13

POSITIVE LOGITS

apus

0.15

 Lover

0.14

lop

0.14

Za

0.13

ãĥŃãĥ³

0.13

ihn

0.13

resizing

0.13

Eug

0.13

 teenager

0.13

iban

0.13

Activations Density 0.017%

references to web links and numeric data

No Comments

No Known Activations