INDEX

Explanations

company names in text

oai_token-act-pair · gpt-3.5-turbo

negations or statements disproving a claim

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

 Palest

-0.71

 photoc

-0.67

çīĪ

-0.65

ilst

-0.62

 indo

-0.62

 sacrific

-0.61

 cryptoc

-0.61

 Fukushima

-0.60

Nik

-0.60

 mainland

-0.59

POSITIVE LOGITS

¬

1.32

Ń

1.27

±

1.20

«

1.18

ĵ

1.16

ĸ

1.15

ĳ

1.14

ª

1.13

£

1.12

Ķ

1.12

Activations Density 0.136%

company names in text

negations or statements disproving a claim

No Comments

No Known Activations