INDEX

Explanations

the acronym "EP" followed by a number

oai_token-act-pair · gpt-3.5-turbo

mentions of specific product lines or models

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ashington

-0.83

bryce

-0.83

bara

-0.79

Kay

-0.76

 marsh

-0.68

 terminals

-0.67

allas

-0.67

bay

-0.67

 milit

-0.66

oldemort

-0.66

POSITIVE LOGITS

EP

4.01

EP

1.79

LP

1.46

Ep

1.39

ep

1.31

Ep

1.27

ep

1.23

TP

1.20

HC

1.14

ECO

1.13

Activations Density 0.023%

the acronym "EP" followed by a number

mentions of specific product lines or models

No Comments

No Known Activations