INDEX

Explanations

references to specific brand names, such as "Oracle" and "Xilinx"

oai_token-act-pair · gpt-3.5-turbo

references to the company Oracle and its stock-related mentions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

irt

-0.84

mble

-0.83

ully

-0.80

ablishment

-0.76

rouse

-0.76

Ģ

-0.76

irting

-0.76

irth

-0.76

llah

-0.75

-0.74

POSITIVE LOGITS

 Oracle

1.27

Oracle

0.99

JD

0.78

 Arena

0.71

 membr

0.69

 Database

0.68

 Woman

0.66

 Tablet

0.66

 Printed

0.66

 Tempest

0.65

Activations Density 0.007%

references to specific brand names, such as "Oracle" and "Xilinx"

references to the company Oracle and its stock-related mentions

No Comments

No Known Activations