INDEX

Explanations

mentions related to products and offerings, particularly those highlighting features or qualities

oai_token-act-pair · gpt-3.5-turbo

positive descriptions of products and brands

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

escal

-0.87

 Tayyip

-0.81

inges

-0.70

Amid

-0.70

 onlook

-0.69

wat

-0.69

walk

-0.67

unch

-0.67

elong

-0.66

asser

-0.66

POSITIVE LOGITS

 specialize

1.04

 specializes

0.99

senal

0.97

 warranty

0.97

 advertise

0.96

 patented

0.96

 recommends

0.90

 marketed

0.88

 logo

0.88

 Warranty

0.86

Activations Density 1.211%

mentions related to products and offerings, particularly those highlighting features or qualities

positive descriptions of products and brands

No Comments

No Known Activations