INDEX

Explanations

references to specific types of vehicles, particularly SUVs and sedans

oai_token-act-pair · gpt-3.5-turbo

mentions of SUVs and sedans

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

bread

-0.77

icial

-0.71

bright

-0.70

ode

-0.69

ãĤ±

-0.67

birth

-0.66

odes

-0.65

itives

-0.65

ttp

-0.65

itored

-0.64

POSITIVE LOGITS

SUV

1.11

 sedan

1.07

illac

0.83

 vans

0.80

 dealership

0.77

 cars

0.73

 caravan

0.72

 shenan

0.72

 parked

0.70

 corrid

0.70

Activations Density 0.007%

references to specific types of vehicles, particularly SUVs and sedans

mentions of SUVs and sedans

No Comments

No Known Activations