INDEX

Explanations

phrases emphasizing honesty

oai_token-act-pair · gpt-3.5-turbo

expressions related to honesty

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

 Krish

-0.76

 Libraries

-0.74

LAN

-0.74

interrupted

-0.72

acid

-0.70

enegger

-0.70

 Autism

-0.69

levels

-0.68

-0.67

Stud

-0.67

POSITIVE LOGITS

 honest

1.08

 truthful

0.92

 honesty

0.85

cipled

0.78

 broker

0.76

onest

0.76

 dece

0.73

parency

0.71

 Honest

0.70

rencies

0.70

Activations Density 0.009%

phrases emphasizing honesty

expressions related to honesty

No Comments

No Known Activations