INDEX

Explanations

references to educational achievements and professional titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

ateria

-0.18

 Female

-0.17

heimer

-0.16

áº¥t

-0.16

Female

-0.16

ibrary

-0.15

indre

-0.15

ONO

-0.15

akest

-0.15

 female

-0.15

POSITIVE LOGITS

 native

0.32

 graduate

0.30

native

0.27

 natives

0.27

 frequent

0.26

 lifelong

0.25

 member

0.25

graduate

0.25

 avid

0.23

 fixture

0.22

Activations Density 0.146%

references to educational achievements and professional titles

No Comments

No Known Activations