INDEX

Explanations

words related to buying or purchasing, specifically focusing on the buyers

oai_token-act-pair · gpt-3.5-turbo

references to buyers in various contexts

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

apo

-0.77

ATIONAL

-0.75

 Cancer

-0.73

 Staff

-0.71

mp

-0.70

 Shell

-0.66

OLOGY

-0.64

ARY

-0.64

 Boss

-0.63

east

-0.63

POSITIVE LOGITS

 buyers

1.33

 buyer

1.14

 beware

0.99

 priced

0.95

 purchasing

0.87

 purchases

0.87

asers

0.87

 horizont

0.85

 sellers

0.85

 buys

0.84

Activations Density 0.010%

words related to buying or purchasing, specifically focusing on the buyers

references to buyers in various contexts

No Comments

No Known Activations