INDEX

Explanations

negative sentiment or criticism towards a specific person or group, particularly in a political context

oai_token-act-pair · gpt-3.5-turbo

negative political sentiments or positions

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

 ILCS

-0.80

ĸļ

-0.80

çīĪ

-0.75

 Twain

-0.71

 Wonderland

-0.69

 layers

-0.68

 partName

-0.68

Nig

-0.68

auga

-0.67

 SHARES

-0.67

POSITIVE LOGITS

social

1.14

government

1.11

choice

1.05

establishment

1.04

immigration

0.99

election

0.98

interest

0.95

fascist

0.94

competitive

0.93

claimed

0.93

Activations Density 0.038%

negative sentiment or criticism towards a specific person or group, particularly in a political context

negative political sentiments or positions

No Comments

No Known Activations