INDEX

Explanations

terms related to measurement or evaluation

oai_token-act-pair · gpt-3.5-turbo

instances of the word "measure" and its variations in various contexts

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

AF

-0.72

aja

-0.68

auri

-0.68

Events

-0.67

 Wend

-0.66

 Airl

-0.64

egg

-0.62

 Trash

-0.60

Ops

-0.59

ente

-0.59

POSITIVE LOGITS

 measure

3.97

 measures

2.27

 Measure

2.22

Measure

2.17

 Measures

2.17

 measurement

1.80

 measuring

1.75

 measured

1.54

 metric

1.46

 estimate

1.39

Activations Density 0.006%

terms related to measurement or evaluation

instances of the word "measure" and its variations in various contexts

No Comments

No Known Activations