INDEX

Explanations

executive job titles and positions

oai_token-act-pair · gpt-3.5-turbo

titles and positions of authority in organizational contexts

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

interstitial

-0.78

pmwiki

-0.67

inventoryQuantity

-0.64

 skirm

-0.64

PLA

-0.63

Request

-0.62

 sidx

-0.60

Option

-0.59

Repeat

-0.59

Fourth

-0.59

POSITIVE LOGITS

 Kevin

1.23

 Geoff

1.22

 Shaun

1.21

 Cheryl

1.20

 Tony

1.19

 Andy

1.18

 Steve

1.17

 Chris

1.17

 Brendan

1.17

 Gareth

1.16

Activations Density 0.155%

executive job titles and positions

titles and positions of authority in organizational contexts

No Comments

No Known Activations