INDEX

Explanations

mentions of guidelines or rules in various contexts

oai_token-act-pair · gpt-3.5-turbo

phrases related to guidelines and regulations

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

joy

-0.86

parts

-0.74

bucks

-0.68

tal

-0.68

Tweet

-0.67

Sus

-0.66

charges

-0.66

 spect

-0.65

ãĥ©ãĥ³

-0.64

some

-0.64

POSITIVE LOGITS

 guidelines

1.14

 governing

1.07

idelines

0.99

 guideline

0.98

 promulg

0.96

 Guidelines

0.90

 prescribed

0.90

 dictate

0.87

 regulating

0.87

 govern

0.87

Activations Density 0.038%

mentions of guidelines or rules in various contexts

phrases related to guidelines and regulations

No Comments

No Known Activations