INDEX

Explanations

studies or research findings

oai_token-act-pair · gpt-3.5-turbo

references to studies and research findings

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

 Leilan

-0.70

 Horus

-0.62

ardless

-0.61

ebus

-0.61

Ago

-0.60

Reply

-0.59

quo

-0.59

Por

-0.58

aples

-0.58

 partic

-0.58

POSITIVE LOGITS

 psychologist

0.87

 study

0.84

 psychologists

0.83

study

0.82

 studies

0.80

Researchers

0.78

 statistic

0.77

uggest

0.76

 estimate

0.74

aez

0.73

Activations Density 0.147%

studies or research findings

references to studies and research findings

No Comments

No Known Activations