INDEX

Explanations

names related to prominent political figures

oai_token-act-pair · gpt-3.5-turbo

proper nouns related to notable individuals, particularly political figures and historical figures

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

atari

-1.04

scape

-0.89

tons

-0.85

tml

-0.85

ansas

-0.82

ris

-0.78

lords

-0.78

manship

-0.77

lan

-0.77

roth

-0.77

POSITIVE LOGITS

 Galile

0.96

 Rossi

0.94

 Nicola

0.84

 Tesla

0.83

zzi

0.81

 Simone

0.79

 Sturgeon

0.79

Äį

0.75

 Nero

0.74

iser

0.74

Activations Density 0.024%

names related to prominent political figures

proper nouns related to notable individuals, particularly political figures and historical figures

No Comments

No Known Activations