INDEX

Explanations

terms related to legal or financial agreements

oai_token-act-pair · gpt-3.5-turbo

references to legal settlements

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

asus

-0.90

ussen

-0.88

inx

-0.83

cemic

-0.80

opes

-0.75

İ

-0.73

enic

-0.73

irl

-0.72

emis

-0.72

swick

-0.71

POSITIVE LOGITS

 settlement

1.15

 settlements

0.98

 agreements

0.95

 Settlement

0.90

 agreement

0.87

 payout

0.80

 Agreement

0.78

 arbitration

0.77

 Payments

0.75

 Transactions

0.74

Activations Density 0.014%

terms related to legal or financial agreements

references to legal settlements

No Comments

No Known Activations