INDEX

Explanations

identifiers related to personal characteristics or identities like race, religion, and sexual orientation

oai_token-act-pair · gpt-3.5-turbo

terms related to sexual orientation, race, and religion

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

Not in Any Lists

No Comments

Negative Logits

 headline

-0.67

akings

-0.67

 bulletin

-0.64

 suites

-0.62

 subsequ

-0.61

 programme

-0.61

 Programme

-0.61

 Regulation

-0.61

prise

-0.60

execute

-0.59

POSITIVE LOGITS

enough

1.10

 enough

1.09

skinned

0.92

 Enough

0.84

ophobic

0.82

 herself

0.82

 oneself

0.80

imately

0.79

 myself

0.78

 anymore

0.77

Activations Density 0.223%

identifiers related to personal characteristics or identities like race, religion, and sexual orientation

terms related to sexual orientation, race, and religion

No Comments

No Known Activations