INDEX

Explanations

technical terms related to compatibility and interoperability

oai_token-act-pair · gpt-3.5-turbo

terms related to compatibility and interoperability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

nan

-0.72

Lie

-0.72

inous

-0.66

 dime

-0.64

 Halifax

-0.64

rue

-0.63

prison

-0.63

 incarcerated

-0.63

Dul

-0.62

ell

-0.61

POSITIVE LOGITS

 compatibility

3.67

 Compatibility

2.36

 interoper

1.94

 incomp

1.87

atibility

1.59

 reliability

1.51

 usability

1.45

 compat

1.45

 durability

1.38

 stability

1.36

Activations Density 0.018%

technical terms related to compatibility and interoperability

terms related to compatibility and interoperability

No Comments

No Known Activations