INDEX

Explanations

references to familial backgrounds and social status

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

ihan

-0.15

/lang

-0.15

cks

-0.14

Moy

-0.13

ilation

-0.13

avs

-0.13

rst

-0.13

rost

-0.13

ocup

-0.13

otlin

-0.13

POSITIVE LOGITS

 backgrounds

0.27

 background

0.27

èĥĮæĻ¯

0.22

 household

0.20

 Background

0.20

 households

0.20

background

0.20

Background

0.19

 mixed

0.19

 environment

0.19

Activations Density 0.082%

references to familial backgrounds and social status

No Comments

No Known Activations