INDEX

Explanations

terms related to fraternities and sororities

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

thal

-0.14

ÃĹ↵↵

-0.13

Ð¾Ð»Ð¾Ñģ

-0.13

Äł

-0.13

sv

-0.13

odb

-0.13

SIM

-0.13

ogy

-0.13

áºŃn

-0.13

 McCart

-0.12

POSITIVE LOGITS

 Sigma

0.37

 chapter

0.37

 Gamma

0.35

Tau

0.35

 Chapter

0.35

 chapters

0.34

 Alpha

0.33

sor

0.33

Pi

0.33

chapter

0.32

Activations Density 0.049%

terms related to fraternities and sororities

No Known Activations