INDEX

Explanations

abbreviations for specific government organizations

oai_token-act-pair · gpt-3.5-turbo

instances of government agency names and their abbreviations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 wrinkles

-0.79

 jelly

-0.76

 lashes

-0.76

 cures

-0.74

 regress

-0.73

 treatments

-0.73

 drift

-0.71

 overd

-0.71

 suicidal

-0.71

 stacking

-0.71

POSITIVE LOGITS

formerly

1.14

USA

1.04

ADA

1.00

whose

1.00

AFP

0.99

NYSE

0.97

Northern

0.94

ESCO

0.94

Canada

0.94

Government

0.93

Activations Density 0.080%

abbreviations for specific government organizations

instances of government agency names and their abbreviations

No Comments

No Known Activations