INDEX

Explanations

terms related to censorship, suppression, and control of information

oai_token-act-pair · gpt-3.5-turbo

terms related to oppression and suppression of individuals or groups

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

coni

-0.79

rapnel

-0.78

oof

-0.77

poral

-0.73

 teasp

-0.72

ukong

-0.71

gars

-0.71

ructose

-0.70

cin

-0.70

coon

-0.70

POSITIVE LOGITS

 legitimate

1.05

 whistleblowers

1.04

 innocent

1.02

 inconvenient

1.01

 lawful

0.96

 freedoms

0.95

 democratic

0.94

 dissent

0.92

 genuine

0.92

 democratically

0.91

Activations Density 0.380%

terms related to censorship, suppression, and control of information

terms related to oppression and suppression of individuals or groups

No Comments

No Known Activations