INDEX

Explanations

terms related to different types of buildings and properties

oai_token-act-pair · gpt-3.5-turbo

references to categories or groups of customers and their characteristics

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

 ðŁĳ

-0.73

answer

-0.72

weet

-0.67

 TAMADRA

-0.66

apter

-0.65

cker

-0.64

feel

-0.64

agonist

-0.63

lash

-0.63

 Paulo

-0.62

POSITIVE LOGITS

 recreational

1.21

 residential

1.18

 outpatient

1.16

 vocational

1.14

 emergencies

1.08

 specialty

1.05

 residences

1.05

 weddings

1.04

commercial

1.02

 ambul

1.00

Activations Density 0.439%

terms related to different types of buildings and properties

references to categories or groups of customers and their characteristics

No Comments

No Known Activations