INDEX

Explanations

sensor-related terms or actions related to detection

oai_token-act-pair · gpt-3.5-turbo

terms related to detection and sensing

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

go

-0.64

 awards

-0.63

 Memorial

-0.61

 halls

-0.60

rights

-0.59

fund

-0.58

ogue

-0.57

 finance

-0.57

stock

-0.57

hug

-0.57

POSITIVE LOGITS

 detected

3.64

 detect

2.12

 detectable

1.97

 detects

1.97

 detection

1.93

 Detected

1.93

 detecting

1.91

 sensed

1.77

 Detection

1.62

 spotted

1.58

Activations Density 0.010%

sensor-related terms or actions related to detection

terms related to detection and sensing

No Comments

No Known Activations