INDEX

Explanations

instances of the word "subsidies."

oai_token-act-pair · gpt-3.5-turbo

references to financial support or assistance programs, specifically subsidies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

uscript

-0.88

angers

-0.73

 Jenner

-0.71

aque

-0.70

 Compos

-0.66

wear

-0.65

 Schmidt

-0.65

press

-0.65

ographed

-0.65

 Alam

-0.64

POSITIVE LOGITS

 subsidies

1.13

 reimburse

1.00

 subsidy

0.99

 rebate

0.94

 payments

0.92

 schemes

0.87

 recip

0.85

 entitle

0.84

 subsid

0.83

fee

0.83

Activations Density 0.025%

instances of the word "subsidies."

references to financial support or assistance programs, specifically subsidies

No Comments

No Known Activations