INDEX

Explanations

prepositions 'in' which are often followed by a location or context

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

awei

-0.66

accompanied

-0.63

 Converted

-0.63

 occasionally

-0.63

 rarely

-0.62

not

-0.61

 intertw

-0.61

 Occasionally

-0.59

 sometimes

-0.59

 respectively

-0.59

POSITIVE LOGITS

 slightest

1.42

any

1.28

nor

1.20

 anymore

1.20

 anybody

1.09

 anything

1.07

ANY

1.07

clusively

0.98

 vain

0.96

 sight

0.96

Activations Density 0.177%

prepositions 'in' which are often followed by a location or context

No Comments

No Known Activations