INDEX

Explanations

terms related to rules or limitations

oai_token-act-pair · gpt-3.5-turbo

phrases related to governmental or legal restrictions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

rious

-0.75

psc

-0.67

past

-0.67

earth

-0.65

enic

-0.64

icus

-0.62

TPP

-0.62

 Generations

-0.62

Asia

-0.61

ousing

-0.61

POSITIVE LOGITS

 imposed

1.15

 restricting

1.08

 restrictions

1.01

 prohibited

0.97

 restricts

0.94

 prohibiting

0.91

 restriction

0.87

 inhib

0.84

 regimes

0.83

 restricted

0.82

Activations Density 0.038%

terms related to rules or limitations

phrases related to governmental or legal restrictions

No Comments

No Known Activations