INDEX

Explanations

phrases related to forgiveness

oai_token-act-pair · gpt-3.5-turbo

terms related to forgiveness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

anto

-0.74

eding

-0.72

ulhu

-0.71

Hunt

-0.70

amen

-0.70

hack

-0.67

Hop

-0.67

rus

-0.67

urg

-0.66

psey

-0.66

POSITIVE LOGITS

 forgiven

1.30

 forgiveness

1.20

 forgive

1.12

 forgiving

0.90

 pard

0.88

give

0.87

 forg

0.83

 pardon

0.83

giving

0.82

ivable

0.81

Activations Density 0.010%

phrases related to forgiveness

terms related to forgiveness

No Comments

No Known Activations