INDEX

Explanations

objects typically found in storage spaces like closets, drawers, shelves, and attics

oai_token-act-pair · gpt-3.5-turbo

references to storage spaces

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

maps

-0.76

param

-0.67

 Archdemon

-0.65

Ath

-0.64

 delegation

-0.64

meric

-0.64

ĪĴ

-0.63

 Hots

-0.61

play

-0.60

 deleg

-0.59

POSITIVE LOGITS

 shelves

1.24

 stocked

1.18

 closet

1.12

 shelf

1.10

 drawer

1.09

 freezer

1.05

 containing

1.03

lid

1.00

 cabinets

0.99

ories

0.96

Activations Density 0.129%

objects typically found in storage spaces like closets, drawers, shelves, and attics

references to storage spaces

No Comments

No Known Activations