INDEX

Explanations

words related to historical or cultural significance

oai_token-act-pair · gpt-3.5-turbo

references to cultural heritage

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

vae

-0.94

kers

-0.78

alties

-0.69

utsu

-0.68

err

-0.67

plet

-0.66

ules

-0.66

secut

-0.65

ching

-0.65

pher

-0.65

POSITIVE LOGITS

 heritage

1.10

conservancy

1.01

 preservation

0.85

artifacts

0.76

 arte

0.73

stones

0.72

 relics

0.71

 corpus

0.70

 legacy

0.68

 richness

0.68

Activations Density 0.010%

words related to historical or cultural significance

references to cultural heritage

No Comments

No Known Activations