INDEX

Explanations

mentions of different precincts and voting patterns in a political context

oai_token-act-pair · gpt-3.5-turbo

references to electoral districts or voting areas

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

*/(

-0.91

————————

-0.84

edom

-0.84

neau

-0.81

gers

-0.80

emort

-0.78

loss

-0.78

ogy

-0.77

pmwiki

-0.77

=-=-=-=-

-0.76

POSITIVE LOGITS

 precincts

1.21

inct

1.03

 precinct

1.03

 stations

0.76

 Prec

0.72

 caucuses

0.70

clave

0.68

 overflow

0.67

 clerk

0.65

 Borough

0.65

Activations Density 0.007%

mentions of different precincts and voting patterns in a political context

references to electoral districts or voting areas

No Comments

No Known Activations