INDEX

Explanations

contractions of words

oai_token-act-pair · gpt-3.5-turbo

negations and expressions of reluctance or dissent

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

API

-0.67

 Thumbnails

-0.63

âĢ¢âĢ¢âĢ¢âĢ¢

-0.62

ument

-0.61

 heterogeneity

-0.57

 Begins

-0.57

 Nieto

-0.55

 Invalid

-0.54

iddles

-0.54

 Xuan

-0.54

POSITIVE LOGITS

 hesitate

1.07

 dare

1.03

 necessarily

0.88

 ordinarily

0.88

 tolerate

0.87

 bother

0.86

 recommend

0.84

 willingly

0.83

 dream

0.83

 normally

0.83

Activations Density 0.092%

contractions of words

negations and expressions of reluctance or dissent

No Comments

No Known Activations

contractions of words

negations and expressions of reluctance or dissent

No Comments

No Known Activations