INDEX

Explanations

mentions of wealth-related terms such as "wealth," "income," and "redistribution."

oai_token-act-pair · gpt-3.5-turbo

discussions about wealth and its distribution

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ãģ¦

-0.66

bsp

-0.64

ATIONS

-0.62

EEE

-0.60

GGGGGGGG

-0.60

thia

-0.60

Jere

-0.59

 Vaughn

-0.59

 Blaz

-0.59

nee

-0.58

POSITIVE LOGITS

 redistribution

1.16

 inequality

1.16

 accumulation

1.11

 disparity

1.00

 holdings

0.98

 amassed

0.98

 inequalities

0.96

 wealth

0.96

 disparities

0.94

 trove

0.93

Activations Density 0.067%

mentions of wealth-related terms such as "wealth," "income," and "redistribution."

discussions about wealth and its distribution

No Known Activations