INDEX

Explanations

references to hotels

oai_token-act-pair · gpt-3.5-turbo

references to hotels

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Matter

-0.73

 Anarchy

-0.72

alez

-0.72

xit

-0.69

 Prompt

-0.68

 Izan

-0.68

ktop

-0.67

cale

-0.67

 advers

-0.65

dit

-0.65

POSITIVE LOGITS

 hotel

1.00

 hotels

0.99

 Hotel

0.96

 accommodations

0.96

 guests

0.94

 rooms

0.89

 resorts

0.89

ibur

0.81

 accommodation

0.80

 lobb

0.78

Activations Density 0.018%

references to hotels

references to hotels

No Comments

No Known Activations