INDEX

Explanations

political figures and related terms

oai_token-act-pair · gpt-3.5-turbo

names of political entities and figures within a context of governance and elections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

sylv

-0.72

 Wilmington

-0.68

 Perkins

-0.68

CFR

-0.67

 Myster

-0.65

 Comet

-0.65

 Daytona

-0.64

 Wonderland

-0.64

 Appalach

-0.63

Weiss

-0.62

POSITIVE LOGITS

jriwal

0.88

 Rohingya

0.88

 Modi

0.87

 Sabha

0.87

 Bhar

0.85

 Hindu

0.82

 Sheikh

0.81

Bh

0.81

 Tamil

0.80

ji

0.80

Activations Density 1.512%

political figures and related terms

names of political entities and figures within a context of governance and elections

No Comments

No Known Activations