INDEX

Explanations

references to figures in a scientific or research context

oai_token-act-pair · gpt-3.5-turbo

references to figures or diagrams within a document

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

role

-0.75

ORE

-0.74

 Zeal

-0.64

die

-0.63

minist

-0.63

soDeliveryDate

-0.62

lopp

-0.61

soType

-0.61

istant

-0.61

 polic

-0.60

POSITIVE LOGITS

Figure

1.04

 depicts

0.87

 illustrates

0.83

 below

0.83

 Figure

0.83

 Exhibit

0.81

 illustrating

0.79

 Caption

0.79

Fig

0.75

 summarizes

0.75

Activations Density 0.029%

references to figures in a scientific or research context

references to figures or diagrams within a document

No Comments

No Known Activations