INDEX

Explanations

names and titles associated with historical figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

ãĤ¤ãĥĦ

-0.16

 Wonderland

-0.15

ollywood

-0.14

uÃ§

-0.14

á»ĩ

-0.13

 ÑĤÐ²Ð°

-0.13

 NÄĽm

-0.13

REDENTIAL

-0.13

çĿ

-0.13

èĬ±

-0.13

POSITIVE LOGITS

 Christian

0.26

 Carl

0.24

 Julius

0.23

 Ferdinand

0.23

Edu

0.23

 Georg

0.23

 August

0.23

 Johann

0.23

 Wilhelm

0.22

 Anton

0.22

Activations Density 0.030%

names and titles associated with historical figures

No Comments

No Known Activations