INDEX

Explanations

references to locations or places, particularly when the word "nearby" is used

oai_token-act-pair · gpt-3.5-turbo

references to locations or entities that are close in proximity

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

olicy

-0.67

 inval

-0.66

starter

-0.64

 Editorial

-0.62

ass

-0.61

 suspensions

-0.59

icz

-0.59

Interview

-0.58

 RELEASE

-0.58

Definition

-0.58

POSITIVE LOGITS

 nearby

3.77

 neighboring

2.23

 neighbouring

2.15

 adjacent

2.03

 adjoining

1.87

 nearest

1.55

 surrounding

1.53

 vicinity

1.26

near

1.22

 downtown

1.19

Activations Density 0.016%

references to locations or places, particularly when the word "nearby" is used

references to locations or entities that are close in proximity

No Comments

No Known Activations