INDEX

Explanations

descriptions related to physical appearance and actions of individuals

oai_token-act-pair · gpt-3.5-turbo

references to body parts and their positions or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ERAL

-0.77

utical

-0.73

actic

-0.71

enhagen

-0.70

 Archdemon

-0.69

 basis

-0.68

gorithm

-0.68

utics

-0.67

Adds

-0.66

 Rewards

-0.64

POSITIVE LOGITS

 locked

1.25

 folded

1.20

 glued

1.15

 tucked

1.15

 covered

1.14

 intact

1.14

 bent

1.13

 tilted

1.13

 pressed

1.12

 blown

1.10

Activations Density 0.168%

descriptions related to physical appearance and actions of individuals

references to body parts and their positions or conditions

No Comments

No Known Activations