INDEX

Explanations

terms related to purging or removal of individuals or things

oai_token-act-pair · gpt-3.5-turbo

instances of the word "purge" and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Trib

-0.73

Hop

-0.71

Hop

-0.69

Rog

-0.68

Barn

-0.68

 Gujar

-0.68

omorph

-0.67

 Ding

-0.64

 Mull

-0.64

 archae

-0.64

POSITIVE LOGITS

 Crusade

1.97

 purge

1.96

 crusade

1.48

idding

1.46

---

1.23

PLA

1.20

 Fiorina

1.16

LCD

1.02

LP

1.02

rete

0.94

Activations Density 0.049%

terms related to purging or removal of individuals or things

instances of the word "purge" and related terms

No Comments

No Known Activations