INDEX

Explanations

terms and concepts related to authority and hierarchy

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

deaux

-0.15

Ð³Ð»

-0.15

 Module

-0.14

 gearing

-0.14

æıĲåĩº

-0.14

 Feinstein

-0.14

ISIBLE

-0.14

oids

-0.13

ÑĪÑĤÐ¾Ð²

-0.13

 tritur

-0.13

POSITIVE LOGITS

 granted

0.52

 awarded

0.45

 accord

0.43

 given

0.42

 conferred

0.38

 Granted

0.36

 bestowed

0.36

Granted

0.35

 grant

0.35

 confer

0.34

Activations Density 0.330%

terms and concepts related to authority and hierarchy

No Comments

No Known Activations