INDEX

Explanations

measurements and quantities mentioned in documents

oai_token-act-pair · gpt-3.5-turbo

measurements and numerical descriptions related to sizes, areas, and distances

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Trend

-0.70

 appropriations

-0.61

 entitlement

-0.57

illard

-0.57

 vouchers

-0.57

Quote

-0.56

bt

-0.55

bid

-0.55

Sales

-0.55

bos

-0.55

POSITIVE LOGITS

 kilomet

1.11

 metre

1.09

1.07

 metres

1.07

 centimeters

1.06

1.05

 kilometres

1.05

1.04

Activations Density 0.150%

measurements and quantities mentioned in documents

measurements and numerical descriptions related to sizes, areas, and distances

No Comments

No Known Activations