INDEX

Explanations

words related to air travel and infrastructure, such as runways, airstrips, and landfills

oai_token-act-pair · gpt-3.5-turbo

references to aviation-related terms and infrastructure

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

å¥³

-0.79

é¾įå

-0.74

Beer

-0.74

arians

-0.71

olver

-0.71

Leader

-0.68

henko

-0.67

onom

-0.65

hiro

-0.64

esson

-0.63

POSITIVE LOGITS

 runway

1.37

lane

0.95

 airst

0.93

 clearance

0.80

selage

0.78

 stalls

0.75

 hangar

0.75

vre

0.75

armac

0.75

 carpet

0.74

Activations Density 0.010%

words related to air travel and infrastructure, such as runways, airstrips, and landfills

references to aviation-related terms and infrastructure

No Comments

No Known Activations