INDEX

Explanations

proper nouns related to a specific name "Pur"

oai_token-act-pair · gpt-3.5-turbo

references to "pur" or variations of the word 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 @ 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

enegger

-1.00

ORGE

-0.80

worthiness

-0.73

 Unch

-0.71

ONES

-0.68

 removable

-0.67

 Werner

-0.65

 diplom

-0.64

 Eisen

-0.63

 Hayden

-0.62

POSITIVE LOGITS

ple

1.12

POSE

1.10

poses

1.10

pose

1.09

ported

1.05

ulent

1.00

pure

0.97

itan

0.93

vey

0.92

cell

0.91

Activations Density 0.010%

proper nouns related to a specific name "Pur"

references to "pur" or variations of the word in various contexts

No Comments

No Known Activations