INDEX

Explanations

instances of the word 'treat' followed by a pronoun or noun

oai_token-act-pair · gpt-3.5-turbo

references to how individuals are treated by others

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 undone

-0.68

 linking

-0.65

marks

-0.62

seeing

-0.61

 prediction

-0.60

 projecting

-0.60

chin

-0.59

 circulation

-0.58

 hindsight

-0.58

undo

-0.57

POSITIVE LOGITS

 differently

1.25

 harshly

1.21

 respectfully

1.17

 unfairly

1.09

 kindly

1.05

 humane

1.02

 unequ

1.00

 disrespect

0.99

 similarly

0.99

 accordingly

0.91

Activations Density 0.147%

instances of the word 'treat' followed by a pronoun or noun

references to how individuals are treated by others

No Comments

No Known Activations

instances of the word 'treat' followed by a pronoun or noun

references to how individuals are treated by others

No Comments

No Known Activations