INDEX

Explanations

phrases related to different types of systems and models

oai_token-act-pair · gpt-3.5-turbo

terms related to voting and subscription models

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãĥ©ãĥ³

-0.88

Sensor

-0.77

Engineers

-0.71

Temperature

-0.69

Destroy

-0.66

é¾įåĸļå£«

-0.66

çĶ°

-0.64

Tea

-0.64

Damage

-0.63

Eva

-0.62

POSITIVE LOGITS

 instituted

0.86

itism

0.83

enfranch

0.81

 abolished

0.80

 legalized

0.79

 whereby

0.76

 allows

0.76

 reinstated

0.75

 incentiv

0.74

 allowed

0.74

Activations Density 0.738%

phrases related to different types of systems and models

terms related to voting and subscription models

No Comments

No Known Activations