INDEX

Explanations

phrases related to politics and legislative actions

oai_token-act-pair · gpt-3.5-turbo

mentions of the House of Representatives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

zsche

-0.79

akeru

-0.68

ilitary

-0.63

 Yiannopoulos

-0.62

 Merlin

-0.62

 Alonso

-0.62

istical

-0.62

ces

-0.61

ISO

-0.61

 Mayer

-0.60

POSITIVE LOGITS

wife

0.93

wives

0.92

House

0.92

 Speaker

0.85

builders

0.84

holder

0.77

 House

0.76

nam

0.75

¬¼

0.74

holders

0.73

Activations Density 0.021%

phrases related to politics and legislative actions

mentions of the House of Representatives

No Comments

No Known Activations