INDEX

Explanations

information about exclusion or negation

oai_token-act-pair · gpt-3.5-turbo

negations and instances of non-existence or absence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

soType

-0.69

_-_

-0.67

issance

-0.65

ORGE

-0.65

çĶŁ

-0.64

 Offline

-0.63

 Puzz

-0.63

assetsadobe

-0.63

roo

-0.62

manship

-0.61

POSITIVE LOGITS

 necessarily

1.10

 otherwise

1.06

 ordinarily

1.00

 normally

0.97

 expressly

0.87

fit

0.87

 traditionally

0.86

 conform

0.85

 belong

0.84

 already

0.83

Activations Density 0.184%

information about exclusion or negation

negations and instances of non-existence or absence

No Comments

No Known Activations