INDEX

Explanations

locations such as warehouses

oai_token-act-pair · gpt-3.5-turbo

references to warehouses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

nir

-0.85

ulus

-0.72

bernatorial

-0.72

azeera

-0.71

pend

-0.70

uyomi

-0.68

utan

-0.68

adic

-0.67

cific

-0.66

uated

-0.65

POSITIVE LOGITS

yard

1.01

 warehouse

0.99

 warehouses

0.97

 Warehouse

0.93

 Depot

0.88

yards

0.81

eers

0.80

 shelves

0.79

raper

0.76

housing

0.75

Activations Density 0.027%

locations such as warehouses

references to warehouses

No Comments

No Known Activations