INDEX

Explanations

terms related to offerings, such as rewards, condolences, critiques, or variations

oai_token-act-pair · gpt-3.5-turbo

instances of offers or rewards

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

_-

-0.68

_.

-0.68

assing

-0.66

 Ernst

-0.65

Ern

-0.63

STER

-0.62

aten

-0.60

 Berger

-0.60

 Verb

-0.59

ï¸ı

-0.59

POSITIVE LOGITS

 condolences

1.01

 congratulations

0.87

 unconditional

0.84

 tantal

0.80

 discounts

0.76

 vouchers

0.76

 amnesty

0.76

 warranties

0.74

ailable

0.72

 plat

0.72

Activations Density 0.147%

terms related to offerings, such as rewards, condolences, critiques, or variations

instances of offers or rewards

No Comments

No Known Activations