INDEX

Explanations

phrases related to moral judgements, particularly the concept of evil

oai_token-act-pair · gpt-3.5-turbo

instances of the word "evil" and its associated contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ribe

-0.78

Lago

-0.78

PsyNetMessage

-0.77

ribes

-0.76

illation

-0.75

UNCH

-0.73

lov

-0.71

aro

-0.71

RESULTS

-0.71

drops

-0.70

POSITIVE LOGITS

 incarn

1.04

 evil

0.96

 mastermind

0.87

 enemy

0.87

 villain

0.86

 genius

0.86

 twin

0.85

 undermin

0.85

 adversary

0.84

 deed

0.83

Activations Density 0.013%

phrases related to moral judgements, particularly the concept of evil

instances of the word "evil" and its associated contexts

No Comments

No Known Activations