INDEX

Explanations

words describing locations and their characteristics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

ocide

-0.80

*/(

-0.74

inctions

-0.73

ubi

-0.72

=-=-=-=-

-0.72

Words

-0.72

ravings

-0.70

kos

-0.70

rists

-0.70

phasis

-0.70

POSITIVE LOGITS

 suburb

1.25

 suburban

1.23

 town

1.21

 rural

1.16

 downtown

1.15

 southwest

1.14

 courtyard

1.14

 northwest

1.13

 neighborhood

1.12

 northwestern

1.12

Activations Density 0.106%

words describing locations and their characteristics

No Comments

No Known Activations