INDEX

Explanations

economic terms related to spending or costs

oai_token-act-pair · gpt-3.5-turbo

terms related to financial expenditures

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

Oracle

-0.83

ger

-0.74

Sie

-0.72

 Ryder

-0.70

 Genie

-0.62

 Marathon

-0.61

 original

-0.61

ce

-0.61

 Cosponsors

-0.61

sell

-0.60

POSITIVE LOGITS

 expenditure

1.38

itures

1.35

 expenditures

1.21

 Expend

1.13

iture

1.03

 incurred

0.89

ments

0.82

ptions

0.82

 expended

0.81

 upkeep

0.81

Activations Density 0.010%

economic terms related to spending or costs

terms related to financial expenditures

No Comments

No Known Activations