INDEX

Explanations

proper nouns related to companies, organizations, and products

oai_token-act-pair · gpt-3.5-turbo

proper nouns, specifically names of companies and brands

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 pinch

-0.70

 curfew

-0.70

 refrain

-0.69

 brute

-0.68

 repetitive

-0.65

 excuse

-0.65

 brigade

-0.64

 deleted

-0.63

liest

-0.63

stood

-0.63

POSITIVE LOGITS

 Holdings

1.22

 Laboratories

1.20

 Ventures

1.18

 Enterprises

1.17

 Labs

1.16

 Worldwide

1.11

 Technologies

1.11

 Corporation

1.08

 Systems

1.07

 Interactive

1.07

Activations Density 0.245%

proper nouns related to companies, organizations, and products

proper nouns, specifically names of companies and brands

No Comments

No Known Activations