INDEX

Explanations

phrases related to leadership or authority

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "ces" and variations thereof

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 judgment

-0.67

Management

-0.66

 trend

-0.65

 transition

-0.65

 treatment

-0.64

 stewards

-0.64

 item

-0.63

 outsourcing

-0.62

OST

-0.60

 attribution

-0.60

POSITIVE LOGITS

ces

1.36

vous

1.01

ippi

0.97

aucas

0.84

uits

0.84

xon

0.82

terday

0.80

lip

0.79

creen

0.79

mond

0.79

Activations Density 0.005%

phrases related to leadership or authority

occurrences of the word "ces" and variations thereof

No Comments

No Known Activations