INDEX

Explanations

references to cameras, especially when they are physically located somewhere or involved in specific actions

oai_token-act-pair · gpt-3.5-turbo

mentions of cameras and related recording devices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Tolkien

-0.70

 Chapter

-0.69

lore

-0.68

ree

-0.67

lling

-0.65

iana

-0.64

supp

-0.63

etic

-0.63

olkien

-0.60

 Gained

-0.60

POSITIVE LOGITS

 cameras

3.95

 camera

2.73

 Camer

2.73

 microphones

2.44

 Camera

2.21

camera

2.08

 photographers

2.00

Camera

1.90

 camer

1.84

 scanners

1.72

Activations Density 0.015%

references to cameras, especially when they are physically located somewhere or involved in specific actions

mentions of cameras and related recording devices

No Comments

No Known Activations