INDEX

Explanations

multipliers or comparisons related to quantities or measurements

oai_token-act-pair · gpt-3.5-turbo

comparative quantities or numerical relationships

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Caption

-0.66

bg

-0.65

VOL

-0.65

IAS

-0.65

statement

-0.64

live

-0.62

alogy

-0.62

activity

-0.62

VICE

-0.62

Connector

-0.61

POSITIVE LOGITS

 hundred

1.61

 thirds

1.60

 thousand

1.44

 dozen

1.34

fold

1.27

teenth

1.26

 quarters

1.23

thirds

1.14

 tenth

1.12

 Hundred

1.08

Activations Density 0.120%

multipliers or comparisons related to quantities or measurements

comparative quantities or numerical relationships

No Comments

No Known Activations